AUTONOMOUS FACADE CLEANING # Understand flight control. Choose sensors with purpose. A knowledge base for flight controllers, perception and precise surface distance, with products, traceable prices and the limits of the technology. ArduPilot · Product platformPX4 · Comparison referenceGlass · Facade · PV[Start budget selection ↗][Understand the 50 cm target →] SURFACE-RELATIVE DISTANCEConcept 45–55 cmdesired nozzle range ≠ Sensor errorMotion & Geometry also matter A ## Maintain nozzle distance Local short-range sensing with a tested error budget. First evaluate ultrasound plus an optical cross-check on a fixed test rig. B ## Understand the surroundings 3D LiDAR, IMU and processing software for maps, obstacles and ego-motion. A LiDAR alone does not fly a cleaning path. C ## Recognise glass Camera images and geometry can identify windows. Transparency, reflections and wet surfaces remain separate test cases. SOFTWARE, HARDWARE AND OPERATION ## What controls what? 02 / 10 PX4 and ArduPilot are flight software. Cube and Pixhawk describe hardware. The ground station plans and displays; fast flight control remains onboard. GROUNDQGroundControlPlanning · Telemetry · Operator↔ ONBOARDCompanion ComputerPerception · SLAM · Paths↔ REAL TIMEArduPilot on Cube / PixhawkEKF · Flight control · Motors Communication uses defined interfaces. Water enable and valve shutdown need their own monitored payload path. This knowledge website is outside the flight-control path. Flight software · Product platform ### ArduPilot / ArduCopter Runs the multicopter on the flight controller: state estimation, attitude, position, motors, flight modes and core failsafes. Selected as the platform for the existing DroneShine product. [Documentation ↗](https://ardupilot.org/copter/docs/introduction.html) Alternative flight software · Comparison ### PX4 An alternative to ArduPilot with documented ROS 2 integration. Useful as a knowledge and research reference. Migration of the established product platform has not been decided. [Documentation ↗](https://docs.px4.io/main/en/ros2/user_guide) Flight controller · Hardware ### Cube / CubePilot Cube is hardware, rather than flight software. The exact module, carrier board, power supply, firmware and board support must be compatible. [Documentation ↗](https://docs.cubepilot.org/user-guides/autopilot/the-cube-module-overview) Hardware standard / Manufacturer ### Pixhawk / Holybro Pixhawk describes standards and hardware families; Holybro builds controllers such as the 6C and 6X. Redundancy and connectors differ by model. [Documentation ↗](https://holybro.com/products/pixhawk-6x) Ground station · Operation ### QGroundControl / Mission Planner Configuration, mission planning, telemetry and log analysis. QGC supports PX4 and ArduPilot; Mission Planner belongs to the ArduPilot ecosystem. Fast distance control runs onboard. [Documentation ↗](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/) Perception and integration ### Companion / ROS 2 / MAVLink An additional computer processes cameras, point clouds, odometry and cleaning plans. ROS 2 is a software environment; MAVLink is a communication protocol. Neither replaces the flight controller. [Documentation ↗](https://ardupilot.org/dev/docs/companion-computers.html) ### Controller & Companion: concrete price references Listed USD shop prices; taxes and shipping are not confirmed as included. The Jetson figure is the manufacturer's MSRP. Exact variants are stated. | Product / Included items | Price | Assessment | [Holybro Pixhawk 6C ↗](https://holybro.com/products/pixhawk-6c)Plastic, without power module | USD 199 | 2 IMUs, 1 barometer; with PM02: USD 217.99. | [Holybro Pixhawk 6X ↗](https://holybro.com/products/pixhawk-6x)Standard v2B with PM02D | USD 422.99 | 3 IMUs, 2 barometers, Ethernet. FC module only: USD 269, out of stock when checked. | [Cube Orange+ ↗](https://irlock.com/products/the-cube-orange-plus)Module only | USD 277 | 3 IMUs; carrier board and other system components required separately. | [Cube Orange+ Standard Set ↗](https://irlock.com/products/cube-orange-plus-standard-set)ADS-B carrier + Power Brick Mini | USD 451 | No GNSS listed in the included items. ADS-B does not detect windows, facades or people. | [Jetson Orin Nano Super ↗](https://developer.nvidia.com/embedded/faq)Developer Kit, manufacturer MSRP | USD 399 | Up to 67 TOPS, 7–25 W. Development computer; plan cooling and power supply separately. Integration: from a measurement to the flight controller - [ArduPilot: Rangefinder and Proximity ↗](https://ardupilot.org/copter/docs/common-rangefinder-landingpage.html) Single-distance and obstacle sensing are different integrations. Check the driver, measurement direction, mode and parameters for the chosen firmware. - [ArduPilot: external odometry ↗](https://ardupilot.org/dev/docs/mavlink-nongps-position-estimation.html) The documentation recommends MAVLink ODOMETRY at a minimum of 4 Hz. Coordinate frames, timestamps, covariances and reset behaviour must match; 4 Hz is not an engineering approval for the cleaning controller. - [PX4: ROS 2 and external position ↗](https://docs.px4.io/main/en/ros/external_position_estimation) Documented companion integration; external vision typically at 30–50 Hz. Check firmware, messages and transforms against the actual version. - [RTK close to facades ↗](https://www.u-blox.com/en/technologies/selective-l5) GNSS multipath and obstruction prevent a general centimetre-level promise near buildings. Surface-relative distance sensing remains a separate task. Supported sensors and protocols do not constitute validated facade control. Check selection and parameters against the exact firmware version. SENSOR CAPABILITIES ## Which capability needs which measurement? 03 / 10 | Sensor family | Possible contribution | Cleaning limitation | Typical integration path | IMU / Barometer / Magnetometer | Estimate attitude, acceleration, altitude or heading | No direct facade distance; consider vibration, magnetic disturbances and pressure effects. | Flight controller / EKF | GNSS / RTK | Global position and waypoints | Buildings cause obstruction and multipath. RTK does not replace local nozzle-distance sensing. | Flight controller + correction data | Ultrasound | Single distance, potentially including transparent surfaces | Angle, membrane wetting, temperature, rotor airflow and echoes influence the measurement. | RS485 / Bridge / Rangefinder driver | Optical single-point sensor | Fast distance along one beam direction | Glass can transmit or reflect the beam; one reading does not define a complete plane. | UART / I²C / Rangefinder | 3D-LiDAR + IMU | Point clouds, planes, obstacles; local odometry / SLAM with software | A point cloud is not yet a finished map. Glass, water, mist and ambiguous geometry remain relevant. | Companion / Time synchronisation / ROS | RGB / Stereo / Depth camera | Recognise windows, frames, materials and surface structure | Semantic recognition does not guarantee metric depth; test reflections, light and droplets. | Companion / Calibrated image processing | mmWave radar | Additional range / motion information; research option when optics degrade | Different resolution and reflection paths; glass / spray suitability depends on the product. No general centimetre guarantee. | Companion / Radar processing | Pressure / Flow / Hose tension | Monitor cleaning state and disturbance forces | Does not measure wall distance. Test the signal path and shutdown behaviour together. | Payload controller / Telemetry Functional assessment based on manufacturer data, autopilot documentation and engineering inference. Any combination needs calibration, measurement-validity monitoring and its own tests. 88Products42Manufacturers17Sensor families MARKET OVERVIEW · 23 OCT 2026 ## Sensors, products & prices 04 / 10 88 international products for distance sensing, mapping, navigation and optical perception. Strengths and weaknesses refer to this particular use case. Currencies are not converted; unconfirmed prices remain “on request”. ### Compare distance sensing and mapping with purpose The selection starts with direct distance sensors. Manufacturer resolution or repeatability is not a guaranteed absolute accuracy. Mounting geometry must distinguish 50 cm sensor distance from 50 cm nozzle distance. 36 distance sensors in the quick table Precision candidates, navigation and height sensors are classified by purpose. The precision-candidate label is an engineering shortlist for your own experiments. Measurement errors apply under the stated conditions. | Product / class | Price | Operating range | Error specification | Assessment at 50 cm sensor distance | [Baumer OADM 20I6480/S14F (10144601)]Precision / short-range candidate | 2,235.80 USDUS shop price; tax and shipping not stated | 100–600 mm | Linearity error ±0.05–2 mm; resolution 15–670 µm. This does not establish an identical total absolute error. | 50 cm is nominally a very good fit; the linearity deviation of up to ±2 mm alone does not establish outdoor flight accuracy. This variant provides no specified distance value above 600 mm. | [Baumer UNDK 30U6103/S14 (10125521)]Precision / short-range candidate | 588.90 EURDistributor price includes 19% VAT; shipping extra | 100–1000 mm; 100 mm blind zone in manufacturer documentation | Repeatability ≤0.5 mm, resolution <0.3 mm; temperature drift ≤2% of target distance. Equivalent overall accuracy is not verified. | 450–550 mm is within the manufacturer range. The 0.5 mm repeatability is not a total error bound; 2% temperature drift calculates to 10 mm at 500 mm. Account for warm-up and older conflicting datasheet revisions. | [DFRobot A01NYUB (SEN0313)]Precision / short-range candidate | 27 USDManufacturer price; check tax, shipping and import charges separately | 280–7500 mm for a flat target with the horn fitted; 280 mm blind zone | Manufacturer: ±10 mm; 1 mm resolution specified separately | 450–550 mm is within nominal range, with the lower target limit only 170 mm beyond the blind zone. The advertised 7.5 m requires the horn; ±10 mm remains a sensor specification under manufacturer conditions. | [DFRobot A02YYUW (SEN0311)]Precision / short-range candidate | 15.90 USDManufacturer price; check tax, shipping and import charges separately | 30–4500 mm; 30 mm blind zone | Manufacturer: ±10 mm; 1 mm resolution is a separate specification | 450–550 mm is well beyond the 30 mm blind zone. The specified ±10 mm leaves theoretical margin under manufacturer conditions, but does not prove ±5 cm nozzle control during spray flight. | [DFRobot Gravity URM09 I²C (SEN0304)]Precision / short-range candidate | 12.90 USDManufacturer price; check tax, shipping and import charges separately | 20–5000 mm; selectable 1.5/3/5 m ranges | Manufacturer lists 1% accuracy and 10 mm resolution; the shop does not explain the percentage reference | 450–550 mm is within range. The advertised 1% and 10 mm output steps do not establish absolute ±5 mm accuracy at 500 mm; treat this as a laboratory candidate. | [DFRobot URM14]Precision / short-range candidate | 99 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. | 0.10–1.50 m | Shop: ±0.3% measurement error; wiki instead states ±0.1%. Clarification required. | 450–550 mm lies within the nominal range. First resolve conflicting manufacturer specifications (shop ±0.3%, wiki ±0.1%) and test rotor airflow, spray water and angled glass on a rig. | [Leuze electronic ODSL 30/D232.01-30M-S12 (50147335)]Precision / short-range candidate | 3,118.08 USDUS shop price; tax and shipping not stated | 0.20–30 m at 6–90% diffuse reflectance; up to 65 m at 90% | ±2 mm at 90% or ±5 mm at 6% diffuse reflectance; observe referencing requirements. Resolution 1 mm; repeatability 2 mm (3σ). | 450–550 mm lies within range. Static ±2/±5 mm looks promising; high weight, referencing, motion lag and glass backscatter may nevertheless rule out drone use. | [LightWare LW20/HA]Precision / short-range candidate | 389 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. | 0.20–100 m | <1 cm at 20–30 °C; <2 cm at 15–20 / 30–35 °C; <5 cm at 0–15 / 35–50 °C; <10 cm at −20–0 / 50–60 °C. | 50 cm is within range; the HA accuracy below 1 cm applies at 20–30 °C. Temperature dependence, glass and sensor geometry can substantially reduce headroom. | [LightWare SF20/HA]Precision / short-range candidate | 359 USDExcluding taxes; available on backorder | 0.20–100 m | <1 cm at 20–30 °C; <2 cm at 15–20 / 30–35 °C; <5 cm at 0–15 / 35–50 °C; <10 cm at −20–0 / 50–60 °C. | 50 cm can be measured. Below 1 cm is only specified at 20–30 °C; at 0–15 °C it is already below 5 cm. Test glass, spray water and dynamic flight errors separately. | [MaxBotix MB1242 I2CXL-MaxSonar-EZ4]Precision / short-range candidate | 34.95 USDManufacturer price; check tax, shipping and import charges separately | 200–7650 mm; reports the nearest detectable target | 1 cm resolution; no verified general absolute error bound. Target size can bias the result. | 450–550 mm is above the 200 mm minimum distance. A 1 cm output step does not establish ±1 cm accuracy; measure temperature and target-size effects on the bench. | [MaxBotix MB7066-100 XL-MaxSonar-WRL1]Precision / short-range candidate | 99.95 USDManufacturer starting price for the standard housing; options, tax and shipping extra | 200–10,680 mm in the full-horn housing; target-dependent | 1 cm serial increments, 2 cm analogue steps; no verified absolute error bound for flight | 450–550 mm is outside the MB7066-100 minimum distance of 200 mm. The family datasheet does not establish general ±1 cm absolute accuracy; do not transfer another MaxBotix model’s 500 mm limit to this SKU. | [MaxBotix MB7389-100 HRXL-MaxSonar-WRMT]Precision / short-range candidate | 109.95 USDManufacturer starting price for standard housing; tax, shipping/import and options extra | 300–5000 mm; below 300 mm typically outputs 300 mm | Typically ±1% or better according to series datasheet; serial resolution 1 mm, analogue resolution 5 mm. Housing and near field affect errors. | The 450–550 mm target band goes below the manufacturer recommendation of 500 mm for highest reliability. Therefore not a first choice as the sole control sensor; consider moving the sensor so its measurement path always remains >500 mm. | [Pepperl+Fuchs UC1000-18GS-IUEP-IO-V15 (304928-100003)]Precision / short-range candidate | 243.20 EURDistributor price excluding VAT and shipping | 70–1000 mm; adjustable evaluation range 90–1000 mm | Characteristic deviation ≤±1% of full scale; repeatability ≤±0.1% of full scale; temperature influence ≤±0.75% of full scale after warm-up | 450–550 mm is within nominal range. ±1% of 1 m means ±10 mm characteristic deviation; also account for the separate temperature influence. The datasheet specifies two minutes of warm-up with enhanced compensation. | [Pepperl+Fuchs UC2000-30GM-IUEP-IO-V15-Y (70137787)]Precision / short-range candidate | Price on requestNo reliable public price verified; quotation required | 90–2000 mm; adjustable evaluation range from 120 mm | Characteristic deviation ≤0.2% of full scale; repeatability ≤0.1% of full scale; temperature influence up to 1.5% of full scale. Resolution ≥0.35 mm. | 450–550 mm is well outside the blind zone. At 2 m full scale, 0.2% calculates to 4 mm, but 1.5% temperature influence is 30 mm; this is not a universal ±4 mm total accuracy. | [Pololu / STMicroelectronics VL53L1X Carrier · Pololu 3415]Precision / short-range candidate | 22.95 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.04–4 m in favourable dark conditions; Short mode typically up to 1.3 m with higher ambient-light immunity. | ST datasheet: typically ±20 mm in darkness and ±25 mm in the defined ambient-light tests; depends on mode, timing budget and target. | 50 cm is generally within range; typical ±20–25 mm mathematically leaves some headroom. A laboratory candidate only: allowance for offset, angle, latency, flight motion and outdoor conditions remains unproven. | [Pololu / STMicroelectronics VL53L4CD Carrier · Pololu 3692]Precision / short-range candidate | 13.95 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 1 mm–1.2 m under favourable conditions; valid range can fall below 50 cm in overcast outdoor light with a dark target. | ST at 33 ms and >200 mm: indoors typically ±3% (white) / ±4% (grey); outdoors overcast typically ±8%. At least 90% of readings within the stated limits. | At 50 cm on a white target, calculated typical error is ±15 mm indoors and about ±40 mm outdoors overcast. Grey 17% targets at 50 cm already exceed the typical range at 90% valid readings. Nozzle control within ±5 cm remains unproven. | [SICK DT35-B15251 (1057652)]Precision / short-range candidate | 701.22 USDUS shop price includes stated tariffs; tax and shipping depend on destination | 0.05–12 m at 90% remission and Speed=Slow; up to 3.1 m at 6% | Typically ±10 mm; resolution 0.1 mm. Repeatability is separate and depends on speed. | 450–550 mm lies within range. Typically ±10 mm does not guarantee ±5 cm nozzle guidance: incidence angle, mounting offset, flight motion and latency add further errors. | [SICK UM18-218126111 (6048404)]Precision / short-range candidate | 326.23 EURRS Netherlands: excluding VAT and shipping | 120–1000 mm operating range; 1300 mm limiting range | ±1% of current measured distance, ±0.15% repeatability; resolution ≥0.069 mm | 450–550 mm is within operating range. ±1% calculates to ±5 mm at 500 mm under manufacturer conditions; this does not prove dynamic ±5 cm performance in wind and spray. | [SICK UM30-213113 (6036918)]Precision / short-range candidate | 1,059.65 USDUS shop price; tax and shipping not stated | 200–1300 mm operating range; limiting range 2000 mm | ±1% of the current measured value; repeatability ±0.15%; resolution ≥0.18 mm | At 500 mm, ±1% calculates to ±5 mm under manufacturer conditions. 450–550 mm lies within the operating range; dynamic nozzle guidance within ±5 cm requires separate verification. | [ifm electronic O1D100 / O1DLF3KG/IO-LINK]Precision / short-range candidate | 570.90 USDUS list price; taxes, shipping and possible duties extra | 0.20–10 m on white reference target; black up to 4 m | At 0.2–1 m/50 Hz: ±15 mm white or ±18 mm grey below 40 klx; ±26.5 mm at 40–100 klx. This is not repeatability. | 50 cm falls within the error specifications for 200–1000 mm. ±15–26.5 mm leaves static allowance relative to the ±50 mm goal; total flight error and measurement dropouts are not established by this. | [microsonic lpc+100/CFI (36310)]Precision / short-range candidate | 301.30 EURFrench distributor price excluding VAT and shipping | 120–1000 mm operating range; 1300 mm limiting range | ±1% with internal temperature compensation; ±0.15% repeatability | 450–550 mm falls within operating range and above the 120 mm blind zone. Temperature-compensated accuracy does not establish performance in rotor airflow and spray. | [microsonic mic+130/IU/TC (22 320)]Precision / short-range candidate | 284.70 EURNet, plus VAT and shipping | 200–1300 mm operating range; limiting range 2000 mm | ±1% with internal temperature compensation; repeatability ±0.15%; resolution 0.18–0.57 mm depending on analogue window | 450–550 mm lies within the operating range and outside the 200 mm blind zone. Distinguish ±1% from resolution and repeatability; no proof of ±5 cm in flight with the nozzle running. | [microsonic pico+100/I]Precision / short-range candidate | 234.10 EURConfirmed net price; taxes and shipping are additional. | 120–1000 mm operating range; limiting range 1300 mm | ±1% with temperature compensation; repeatability ±0.15%; resolution 0.069–0.38 mm depending on analogue window | 450–550 mm lies well within the operating range. ±1% remains a sensor specification under measurement conditions; verify the entire nozzle distance band with motion, spray water and signal dropouts. | [microsonic zws-70/CI/QS]Precision / short-range candidate | 234.10 EURConfirmed net price; taxes and shipping are additional. | 120–700 mm operating range; 1000 mm limiting range | ±1% with internal temperature compensation; ±0.15% repeatability, 0.20 mm resolution | 450–550 mm is well beyond the 120 mm blind zone and within operating range. The ±1% sensor specification and 0.20 mm resolution do not establish guaranteed flight accuracy. | [Benewake TF-Luna · 8 m (DFRobot SEN0340)]Range / navigation | 24.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.20–8 m at 90% reflectance; only up to 2.5 m at 10% reflectance (indoors). | ±6 cm at 0.2–3 m; ±2% at 3–8 m. 1 cm resolution is not an accuracy specification. | 50 cm lies within range, but ±6 cm at 50 cm does not meet ±5 cm nozzle distance. Sensor offset, tilt, latency and flight motion add further errors. | [Benewake TF02-Pro · UART/I²C (DFRobot SEN0238)]Range / navigation | 87.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.10–40 m at 90% reflectance; 0.10–13.5 m at 10%, including 100 klux. | ±5 cm at 0.1–5 m; ±1% at 5–40 m; reference: white surface, 25 °C. | 50 cm is within nominal range, but ±5 cm already consumes the entire target budget. A mounting offset increasing sensor distance does not change this near-range limitation. | [Benewake TF02-i · RS-485 (DFRobot SEN0504)]Range / navigation | 109.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.10–40 m at 90% reflectance; up to 13.5 m at 10%, including 100 klux. | Manufacturer A03: ±5 cm at 0.1–5 m; ±1% at 5–25 m. No accuracy promise for 25–40 m in that document. | At 50 cm, nominal error is already ±5 cm. Offset, angle, latency and flight motion require additional allowance; no assurance of 50 cm ±5 cm. | [Benewake TF03-100 Standard UART/CAN]Range / navigation | 229.11 USDRobotShop US price includes US tariffs according to listing; VAT/sales-tax status not explicitly stated, shipping extra. | 0.1–100 m at 90% reflectance; 40 m at 10%; at 100 klux 80 m and 30 m respectively according to A01 | ±10 cm below 10 m; ±1% from 10 m; 1 cm resolution is not accuracy | At 50 cm, ±10 cm applies according to A01 datasheet. Insufficient for the ±5 cm total goal without an additional, more precise measurement channel. | [Benewake TFS20-L · 20 m OEM module (DFRobot SEN0672)]Range / navigation | 43.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.20–20 m at 90% reflectance in darkness; 15 m at 100 klux. Dark target: 12 m indoors / 9 m at 100 klux. | ±6 cm at 0.2–6 m; ±1% from 6 m. Repeatability 2 cm (1σ) is separate. | 50 cm lies within range; ±6 cm absolute accuracy exceeds the total goal. 2 cm repeatability does not replace absolute error and does not establish flight-control performance. | [Benewake TFmini Plus]Range / navigation | 49.90 USDTax status not stated | 0.10–12 m; dark targets up to 4 m | ±5 cm at 0.1–5 m; ±1% at 5–12 m according to A07 datasheet. | 50 cm lies within range, but ±5 cm sensor error consumes the entire permissible nozzle deviation. No allowance remains when used as the sole control sensor for this goal. | [Garmin LIDAR-Lite v4 LED · 010-02022-00]Range / navigation | 69.99 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. | 0.05–10 m; reference point according to retailer: rear of the device. | Repeatability in High Accuracy mode indoors on a 90% target: ±1 cm up to 2 m, ±2 cm up to 4 m, ±5 cm up to 10 m (1σ). No general absolute-error guarantee. | 50 cm lies within range and ±1 cm repeatability sounds promising. Absolute accuracy, sunlight, wet glass, offset and latency still lack complete verification for ±5 cm. | [LightWare LW20/C · 100 m (standard, not HA)]Range / navigation | 349 USDManufacturer price explicitly excludes taxes; shipping and import extra. | 0.20–100 m; 70% albedo in sunlight, 0.9 × 0.9 m target. | ±5 cm below 500 readings/s; ±10 cm above 500/s. Do not apply HA accuracy. | At 50 cm, ±5 cm at low rate consumes the entire target budget; high rate is specified at ±10 cm. Not verified for ±5 cm total nozzle standoff. | [LightWare SF000/B · 50 m]Range / navigation | 319 USDManufacturer price explicitly excludes taxes; shipping and import extra. | 0.20–50 m; 70% albedo in sunlight, 0.9 × 0.9 m target. | ±5 cm according to product guide 7.3; 1 cm resolution. | 50 cm can be measured, but ±5 cm sensor error leaves no allowance. Calibrate nozzle mounting offset and time delay on the complete system. | [Ainstein US-D1 Radar Altimeter]Height sensing | 599 USDTax status not explicitly stated in manufacturer shop; shipping extra. | 0.5–50 m; blind zone below 0.5 m | Manufacturer precision 6 cm below 1 m, 4 cm above 1 m; no demonstrated control accuracy | 50 cm is the lower measurement limit. The 450–550 mm target band intersects the blind zone; precision is also 6 cm below 1 m. Classify as an altimeter, not a precise facade-control sensor. | [LightWare SF30/D · 200 m]Height sensing | 459 USDManufacturer price explicitly excludes taxes; shipping and import extra. | 0.20–200 m; 70% albedo in sunlight, 10 km visibility, 0.9 × 0.9 m target. | ±5 cm below 500 readings/s; ±10 cm above 500 readings/s. The boundary at exactly 500/s is unclear in the guide. | 50 cm lies within range; ±5 cm at low or ±10 cm at high rate is insufficient as overall verification of 50 cm ±5 cm. Primarily classify as a height/backup channel. | [Nanoradar NRA24 (current 200 m version)]Height sensing | Quote requiredCurrent 200 m configuration: quote required. Prices for older 50 m versions were not treated as the same product. | 0.5–200 m according to the current manufacturer page | Current manufacturer specification: ±0.1 m. Do not transfer older 50 m documents claiming 2 cm to the 200 m variant. | Additional height measurement only: current ±10 cm and 0.5 m minimum range do not meet the 45–55 cm facade target. 28 mapping devices in the quick table 2D scanners produce a scan plane; a spatial model requires additional motion and the corresponding pose. 3D LiDAR produces a point cloud; registration or SLAM and calibration turn it into a map. A survey payload may integrate additional IMU / GNSS components and must be assessed separately from a bare sensor. | Product / mapping | Price | Operating range | Weight / power | Required mapping pipeline | [Hokuyo UST-20LX]2D scan plane | 2,758.80 EURRobotShop EU, SKU RB-Seu-11, including taxes; shipping extra. | 0.06–20 m on white Kent paper; up to 8 m at 10% reflectivity; 60 m is only the maximum detection distance | 130 g without cable10–30 V; ≤150 mA at 24 V (≤3.6 W), startup about 450 mA | 270° contour in one plane. SLAM and motion create larger 2D maps; 3D requires measured motion/orientation. Not a complete flight system. | [LDROBOT STL-27L – Waveshare Kit 24782]2D scan plane | 167.99 USDWaveshare kit price SKU 24782; check tax/import status and shipping separately. | 0.03–25 m at about 80% reflectivity | 46.1 g sensor according to Waveshare; kit heavier5 V, about 290 mA / 1.45 W | Planar distance scan. Continuous 2D maps require SLAM and motion; 3D additionally requires known scan-plane orientation/motion. Not a complete navigation system. | [LightWare SF45/B (50 m)]2D scan plane | 519 USDManufacturer shop explicitly excludes taxes; shipping and import duties extra. | 0.2–50 m; maximum range specified with a 70% reflective target | 59 g without cable according to current guide; shop text additionally gives older 58 g figure5–5.5 V; 300 mA typical, <450 mA at startup | Planar scan: a continuous 2D map requires motion and poses/SLAM; a spatial map additionally requires known scan-plane motion and orientation. | [SICK TiM571-2050101 (1075091)]2D scan plane | 2,015.96 EURRS Germany guide price for 1 unit, excluding VAT; €2,398.99 including VAT; shipping separate. | 0.05–25 m; 8 m at 10% remission | 250 g without cable9–28 V DC; typically 4 W | Provides a planar 270° scan. Contour maps result from registration using SLAM/poses. Spatial maps require motion and known scan-plane orientation. | [SLAMTEC RPLIDAR A3 (A3M1)]2D scan plane | 599 USDDFRobot unit price, SKU DFR0583; check tax/import status and shipping separately. | From 0.2 m; up to 25 m on white targets in Enhanced mode, 20 m in Outdoor mode | 190 g5 V; 450–600 mA, 2.25–3 W | Planar contour map after registering multiple scans with estimated poses. Spatial reconstruction requires controlled scan-plane motion and registration with IMU/poses. | [SLAMTEC RPLIDAR S2 (S2M1-R2)]2D scan plane | 399 USDSeeed unit price, explicitly excluding VAT; shipping and import extra. | 0.05–30 m at 90%; up to 10 m at 10% reflectivity | 185 g according to manufacturer datasheet; some retailers state 190 g5 V; manufacturer states >2 W; account for startup current | Produces planar distance contours. A continuous map requires motion and estimated poses/SLAM; a 3D map additionally requires motion with a precise orientation reference. No factory-provided autonomous navigation. | [SLAMTEC RPLIDAR S3]2D scan plane | 549 USDSeeed unit price; VAT status unclear on product page, check shipping/import separately. | 0.05–40 m at 70%; 15 m at 10%; 5 m at 2% reflectivity | 115 g5 V, typically 450 mA (about 2.25 W); startup typically 1.2 A | Provides planar 360° points. A larger 2D map requires motion and SLAM/poses; 3D additionally needs known motion and orientation. Not complete flight navigation. | [YDLIDAR / EAI TG30]2D scan plane | 359 EURExp-Tech SKU 750-TG30, explicitly excluding VAT and plus shipping; B2B shop. | 0.05–30 m; account for target reflectivity and lighting | 140 g net according to manufacturer datasheet V1.4; older retailer documents state 214 g5 V (4.8–5.2 V); typically 340 mA, about 1.7 W; startup up to 1 A | Individual planar scans form a registered map only with motion and poses. 3D reconstruction requires additional known scan-plane motion and orientation. | [Hesai JT128]3D point cloud | 1,200 EURNet, excluding 19% VAT and shipping according to offer. | 0–60 m instrumented range; 40 m at 10% reflectance across all channels. | Under 265 g according to current manual; product page rounds to 265 g.8.4 W typical at 25 °C and 10 Hz, without accessories, according to directly reviewed manufacturer manual J01-en-261010 (manual pp. 17–18, PDF pp. 21–22); 9–32 V DC. Allow at least 30 W available supply. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Hesai JT16]3D point cloud | 479 EURIncludes statutory German VAT according to retailer tax section; product page simultaneously shows a conflicting additional tax note. Confirm at checkout. | 0.15–100 m instrumented measurement range; 30 m at 10% reflectance across all channels, 50% detection probability and 0–100 klx. | 199.7 g.4.3 W typical at 14 V / 5 Hz; 12–16 V DC; at least 16 W supply headroom according to manual. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Hesai XT16]3D point cloud | Price on requestRequest a current quotation | 0.05–120 m | 800 gNo reliable figure in the reviewed excerpt; check manufacturer datasheet before design | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Hesai XT32]3D point cloud | Price on requestNo reliable public new-unit price verified | 0.05–120 m | 800 gNo reliable figure in the reviewed excerpt; check manufacturer datasheet before design | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Hesai XT32M2X (XT32M family)]3D point cloud | 4,636.40 EURNet, excluding VAT; shipping and project-specific accessories extra. | 0.5–300 m measurement range; 80 m at 10% reflectance across all channels (100 klx, >90% detection probability). | 490 g without additional integration accessories.10 W typical; 9–36 V DC; manufacturer requires at least 30 W available supply power. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Livox Avia]3D point cloud | 1,420 EURGebrüder Kraus: net; €1,689.80 including 19% VAT, shipping from €25. | At 100 klux: 190 m at 10%, 320 m at 80%; 450 m at 80% only at 0 klux | 498 g without cable8 W non-repetitive / 9 W repetitive, startup 16 W; 10–15 V (Converter 2.0: 9–30 V) | Produces actual spatial points. Continuous maps require LiDAR/IMU SLAM and poses; georeferenced survey results additionally need GNSS/INS and calibrated processing. Not a complete payload. | [Livox MID-360]3D point cloud | Not orderable from TPINo current purchase price from the reviewed retailer | 0.10–70 m, target dependent | 265 g6.5 W average; 9–27 V | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Livox MID-360S]3D point cloud | 599 EURConfirmed net price; taxes and shipping are additional. | 0.10–100 m; 40 m at 10% reflectance | 265 g6.5 W; 9–27 V DC | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Ouster OS0-128 Rev8]3D point cloud | 14,923 EURIncludes 19% VAT; shipping extra; special order | 35 m at 10% and 75 m at 80% reflectance (100 klx, >90% detection probability); 500 m is the encodable range. Minimum distance 0.5 m default, optional 0.3/0.0 m. | 500 g without cap/baseplate; 580 g with cap.15 W nominal, 10–20 W at 1024 × 10 Hz; supply at least 30 W. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Ouster OS1 Max Rev8 (64 / 128 / 256 channels)]3D point cloud | Price on requestPrice on request; taxes and shipping depend on quotation. | 200 m at 10% and 350 m at 80% reflectance (100 klx, >90% detection probability); 500 m maximum encodable range. Minimum range 0.5 m default; optional 0.3/0.0 m. | 670 g without cap/baseplate; 720 g with cap.19 W nominal, 15–25 W at 1024 × 10 Hz; allow at least 30 W supply power. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Ouster OS1 Rev8 (32 / 64 / 128 channels)]3D point cloud | Price on requestPrice on request; taxes and shipping depend on quotation. | 90 m at 10% and 170 m at 80% reflectance (100 klx, >90% detection probability); 500 m maximum encodable range. Minimum range 0.5 m default; optional 0.3 or 0.0 m. | 500 g without cap/baseplate; 580 g with cap.15 W nominal, 10–20 W at 1024 × 10 Hz; size supply for at least 30 W. | Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. | [Ouster OSDome Rev8 (32 / 64 / 128 channels)]3D point cloud | Price on requestPrice on request; taxes and shipping depend on the quotation. | 20 m at 10% and 45 m at 80% reflectivity (100 klx, >90% detection probability). Datasheet: 500 m encodable; minimum range 0.5 m standard, optionally 0.3/0.0 m. | 540 g without baseplate.15 W nominal; 10–20 W, plus allowance for accessories and startup. | Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. | [RoboSense Airy · 96 Beam]3D point cloud | 999 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. | 0.10–60 m; 30 m at 10% reflectivity | 230 ±20 g8 W typical; 9–32 V DC | Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. | [RoboSense Fairy 48-Beam – Device Only]3D point cloud | 1,699 USDManufacturer shop mentions DDP and duties/import taxes calculated at checkout. Check the destination-specific final amount; accessory kit is separate. | Up to 150 m manufacturer maximum; the reviewed product page does not separately establish minimum detection distance. 0.5 m is the lower limit of a published precision test, not necessarily the blind zone. | Under 350 g according to manufacturer; φ75 × 70 mm.No verified power figure for the exact 48-channel variant on the reviewed manufacturer pages; request a binding datasheet. | Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. | [RoboSense Helios 32 (H32F70)]3D point cloud | Price on requestExact H32F70 unit price is unclear in the multi-variant shop; request a Device Only quotation. | Up to 150 m; 110 m at 10% NIST reflectivity; blind zone ≤0.2 m | 1,000 ±50 g9–32 V; 12 W at 10 Hz | Produces spatial surround point clouds. Mapping needs a trajectory/poses, IMU, synchronization and a SLAM computer; georeferencing also needs GNSS/INS. Not a complete flight-navigation package. | [RoboSense RS-Bpearl (offered 0.92 kg configuration)]3D point cloud | 2,436 EURTax status is unclear on the reviewed product page; confirm the final price and shipping. | For this documented configuration: 0.1–30 m at 10% reflectivity; manufacturer specifies a near blind zone ≤10 cm. | 0.92 kg without cable for this listing.13 W typical, 9–32 V DC in associated manufacturer documentation. | Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. | [Seyond Robin E1X (192-line configuration)]3D point cloud | 3,750 USDCRATUS sales terms: USD excluding taxes, duties, shipping, insurance and fees; other DDP quotations may differ. | 1–250 m according to the current manufacturer page and 192-line brochure; 200 m at 10% reflectivity. Older 96-line documents specify up to 300 m. | 470 g.6 W according to associated manufacturer brochure; 9–16 V DC. | Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. | [RIEGL VUX-1UAV²² — scanner only]UAV survey | Price on requestManufacturer quote depends on GNSS/INS, software and integration; no verified public fixed price. | PRR-dependent: at 50 kHz up to 755 m @20% / 1,415 m @80%; at 1,200 kHz 160 / 320 m; minimum distance 1.5–3 m | 3.5 kg without / 3.75 kg with fan; excluding external GNSS/IMU11–34 V; typically 60 W excluding external GNSS/IMU | The line scanner produces a georeferenced 3D map only through flight motion and precise GNSS/IMU poses. Do not confuse a bare scanner with a complete VUX-SYS payload; not a cleaning-navigation system. | [RIEGL miniVUX-3UAV — scanner only]UAV survey | Price on requestConfiguration and price on manufacturer request; no publicly verified fixed unit price. | From 2 m; up to 170 m at ≥20% / 330 m at ≥80% reflectivity (mode-dependent) | 1.55 kg without / 1.60 kg with fan; excluding external GNSS/IMU11–34 V; typically 18 W at 100 scans/s | A moving line scanner combined with a GNSS/IMU trajectory produces a georeferenced 3D point cloud. Distinguish the bare scanner from an integrated miniVUX-SYS/RiLOC payload; no autonomous cleaning navigation. | [YellowScan Mapper+]UAV survey | Price on requestManufacturer quote depends on mount, software and accessories; no verified public fixed price. EUR/USD 0 shop placeholders are not prices. | Recommended maximum flight altitude 100 m AGL; laser range up to 230 m | 1.1 kg without battery; manufacturer lists 1.3 kg for the supplied Mapper+35 W; own battery or power through the platform interface | Integrated LiDAR/GNSS/IMU payload for a georeferenced 3D point cloud after trajectory and point-cloud processing. Includes more than a bare sensor, but does not replace autonomous cleaning or collision control. Product lifecycle: excluded legacy products - [Excluded: SF11/C is end of life and sold out ↗](https://lightwarelidar.com/shop/sf11-c-100-m2/) · Manufacturer states end of life on 31 October 2024. The displayed US$199 price is not an available new-product offer. - [Excluded: Terabee Evo Mini / Evo 60m ↗](https://www.mouser.com/PCN/Terabee_Terabee_EOL_Letter_Evo_range___Mouser_Electronics.pdf) · Manufacturer letter dated 23 May 2024 discontinues the entire distance-sensor range with immediate effect; do not base a new development project on it. Ouster Rev8: understanding models and updates The current Ouster Rev8 family includes OS0, OS1, OS1 Max and OSDome. Models differ in range, field of view, channel count and integration budget. Manufacturer documentation lists OS2 as Rev7. ### What improves - Native RGB-D colour on the sensor; colour and distance are generated together. - OS0/OS1 claim twice the precision and accuracy of Rev7; up to 4096 horizontal samples and up to 40 Hz depending on mode. - OS1 Max adds up to 256 channels and 200 m detection on 10% targets; robust housings and planned functional-safety features. ### What still needs testing - The claim of doubled range/resolution describes the family; specific models, scan modes and target reflectance remain decisive. - An adjustable 0.0 m minimum distance does not guarantee accuracy at 0.5 m. Accuracy/precision testing in these datasheets applies beyond 1 m. - 1 mm distance resolution, 1σ precision and ± accuracy are separate quantities. - IP68/IP69K describes housing protection with cable connected; it does not establish measurement through spray mist, droplets or on glass. - Designed for functional safety does not establish completed certification or the safety of an entire cleaning drone. - Measurement of sensor-to-nozzle offset, tilt, motion and control determine whether 50 cm ±5 cm is achieved; no datasheet guarantees it. [Ouster: current Rev8 product family ↗](https://ouster.com/rev8) · [Ouster OS0 Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os0.pdf) · [Ouster OS1 Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os1.pdf) · [Ouster OS1 Max Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os1-max.pdf) · [Ouster OSDome Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-osdome.pdf) · [Ouster: OS2 explicitly Rev7 ↗](https://static.ouster.dev/sensor-docs/hw_user_manual_OS2/hw_common_sections_OS2/os2-overview.html) Price comparisons need context. Net, gross and USD prices are identified. Taxes, shipping, cables, power supply, computer, mounts and integration may add cost. Earlier chat prices have been replaced with newly checked figures. Direct distance sensing: precision candidates, navigation and height sensors. 88 productsSelect up to 3 products for direct comparison Laser triangulation sensorCompare: Baumer OADM 20I6480/S14F (10144601) Baumer ### OADM 20I6480/S14F (10144601) Very precise local distance, preferably on a test rig or in controlled lighting 2,235.80 USDUS shop price; tax and shipping not stated High geometric precision around 50 cm, but little ambient-light headroom for sunlight. Assessment at 50 cm sensor distance: 50 cm is nominally a very good fit; the linearity deviation of up to ±2 mm alone does not establish outdoor flight accuracy. This variant provides no specified distance value above 600 mm. Surface distance #### Strengths - Measurement range covers the entire 450–550 mm band. - Small light spot and short response time support precise test-rig experiments. #### Weaknesses - Ambient-light immunity <10 klx is unfavourable in direct sunlight. - Only 50 mm of headroom between the upper 550 mm target limit and the end of the measurement range. IntegrationHigh Regulated 24 V supply, precise analogue input and a custom measurement bridge; scale the output window to the working range and detect out-of-range readings. Operating range / quantity 100–600 mm Error / dispersion Linearity error ±0.05–2 mm; resolution 15–670 µm. This does not establish an identical total absolute error. Interface 4–20 mA and 0–10 V; PNP alarm; M12, 8-pin Specifications, availability & sources Rate / points Response time <0.9 ms Weight / housing No weight on manufacturer page; retailer lists 0.50 lb (about 227 g), net/shipping weight unclear · IP67; Class 2 laser Supply / power 12–28 V DC, max. 100 mA without load Field of view / coverage Spot beam with 2 mm diameter Wet glossy surfaces, transparent surfaces and daylight require testing in the actual setup. Linearity is not a complete error budget. Availability when checked Exact variant listed by manufacturer; public Wolf Automation price, stock and lead time unclear Purchase / region Baumer international; price source USA - [Baumer: OADM 20I6480/S14F technical data ↗](https://www.baumer.com/us/en/product-overview/distance-measurement/laser-distance-sensors/oadm-20i6480-s14f/p/26307) - [Wolf Automation: OADM 20I6480/S14F price and retailer weight ↗](https://www.wolfautomation.com/laser-distance-sensor-44396) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.wolfautomation.com/laser-distance-sensor-44396) UltrasonicCompare: Baumer UNDK 30U6103/S14 (10125521) Baumer ### UNDK 30U6103/S14 (10125521) Rectangular ultrasound with analogue output 588.90 EURDistributor price includes 19% VAT; shipping extra Alternative form factor for a fixed bench setup; account for its long warm-up period. Assessment at 50 cm sensor distance: 450–550 mm is within the manufacturer range. The 0.5 mm repeatability is not a total error bound; 2% temperature drift calculates to 10 mm at 500 mm. Account for warm-up and older conflicting datasheet revisions. Surface distance #### Strengths - Teach-in and a straightforward analogue distance signal. - The 100 mm blind zone leaves geometric margin for 50 cm. #### Weaknesses - Manufacturer specifies approximately 15 minutes to compensate power-up drift. - 0.5 mm repeatability must not be represented as absolute accuracy. IntegrationHigh Requires regulated 24 V, a 0–10 V to 3.3 V measurement adapter/ADC and a rangefinder bridge. Measure scaling, warm-up drift and actual response on the assembly. Operating range / quantity 100–1000 mm; 100 mm blind zone in manufacturer documentation Error / dispersion Repeatability ≤0.5 mm, resolution <0.3 mm; temperature drift ≤2% of target distance. Equivalent overall accuracy is not verified. Interface 0–10 V / 10–0 V, M12; teach-in Specifications, availability & sources Rate / points <80 ms response/release time Weight / housing Net weight not verified · IP67 Supply / power 15–30 V DC, <35 mA without load Field of view / coverage Manufacturer beam diagram; no verified universal angle Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Exact /S14 variant (10125521) listed with a public distributor price; confirm lead time Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [Baumer: current technical specifications for the exact /S14 variant ↗](https://www.baumer.com/int/en/p/33867) - [Baumer: /S14 manual, range and power-up drift ↗](https://media.baumer.com/m_undk30u6103s14.pdf) - [Elektro4000: exact order number and VAT-inclusive price ↗](https://www.elektro4000.de/Steuerungen-Schaltgeraete/Sensoren/Ultraschall/Abstandssensoren/Baumer-Sensor-UNDK-30U6103-S14%3A%3A1001947.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.elektro4000.de/Steuerungen-Schaltgeraete/Sensoren/Ultraschall/Abstandssensoren/Baumer-Sensor-UNDK-30U6103-S14%3A%3A1001947.html) UltrasonicCompare: DFRobot A01NYUB (SEN0313) DFRobot ### A01NYUB (SEN0313) Enclosed ultrasonic sensor with longer range 27 USDManufacturer price; check tax, shipping and import charges separately Affordable longer-range comparison; account for the horn and nearby edges in the test setup. Assessment at 50 cm sensor distance: 450–550 mm is within nominal range, with the lower target limit only 170 mm beyond the blind zone. The advertised 7.5 m requires the horn; ±10 mm remains a sensor specification under manufacturer conditions. Surface distance #### Strengths - Enclosed sensing head with a public manufacturer price. - Digital output and low supply voltage. #### Weaknesses - The 280 mm blind zone leaves less near-range margin than the A02YYUW. - A broad beam and 100 ms response complicate fast precise façade tracking. IntegrationMedium Use 3.3 V operation or verified level conversion and a UART bridge; pass data age and invalid-echo status to the flight controller. Calibrate the sensor-to-nozzle offset. Operating range / quantity 280–7500 mm for a flat target with the horn fitted; 280 mm blind zone Error / dispersion Manufacturer: ±10 mm; 1 mm resolution specified separately Interface UART, 9600 baud Specifications, availability & sources Rate / points 100 ms response time Weight / housing Not verified · IP67 Supply / power 3.3–5 V DC, <15 mA average Field of view / coverage 55° with horn, 40° without horn; range depends on assembly Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Listed at USD 27 in the manufacturer shop; confirm stock and horn configuration Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [DFRobot: A01NYUB specifications, horn and price ↗](https://www.dfrobot.com/product-1934.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1934.html) UltrasonicCompare: DFRobot A02YYUW (SEN0311) DFRobot ### A02YYUW (SEN0311) Low-cost enclosed near-range test 15.90 USDManufacturer price; check tax, shipping and import charges separately Affordable candidate for dry and wet glass experiments; account for its broad beam. Assessment at 50 cm sensor distance: 450–550 mm is well beyond the 30 mm blind zone. The specified ±10 mm leaves theoretical margin under manufacturer conditions, but does not prove ±5 cm nozzle control during spray flight. Surface distance #### Strengths - Short blind zone and low public price. - Enclosed construction and UART simplify an initial bench setup. #### Weaknesses - The 60° reference angle may include frames, the boom or adjacent surfaces. - A 100 ms response limits useful control-loop bandwidth. IntegrationMedium Use a 3.3 V supply or suitable signal-level conversion; read UART data through a verified rangefinder driver or custom bridge. No native MAVLink/CAN output. Operating range / quantity 30–4500 mm; 30 mm blind zone Error / dispersion Manufacturer: ±10 mm; 1 mm resolution is a separate specification Interface UART, 9600 baud Specifications, availability & sources Rate / points 100 ms response time Weight / housing Not verified · IP67; assess the connector and installation separately Supply / power 3.3–5 V DC, ≤8 mA average Field of view / coverage 60° reference angle; target-dependent acoustic beam Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Listed at USD 15.90 in the manufacturer shop; confirm stock before ordering Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [DFRobot: A02YYUW specifications, ±1 cm and price ↗](https://www.dfrobot.com/product-1935.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1935.html) UltrasonicCompare: DFRobot Gravity URM09 I²C (SEN0304) DFRobot ### Gravity URM09 I²C (SEN0304) Fast low-cost laboratory ultrasound 12.90 USDManufacturer price; check tax, shipping and import charges separately Very inexpensive dry bench platform for measurement timing and software; added covers may alter acoustics. Assessment at 50 cm sensor distance: 450–550 mm is within range. The advertised 1% and 10 mm output steps do not establish absolute ±5 mm accuracy at 500 mm; treat this as a laboratory candidate. Surface distance #### Strengths - Configurable range, temperature compensation and up to 50 Hz. - 3.3 V-compatible development setup with a low purchase price. #### Weaknesses - Exposed transducers and PCB are unsuitable for direct spray exposure. - 10 mm resolution and an unclear percentage reference do not establish a ±5 mm accuracy guarantee. IntegrationMedium Keep I²C wires short and match pull-ups to the flight controller’s 3.3 V logic; verify a driver for the exact firmware version or build a bridge. Do not confuse it with the SEN0307 analogue variant. Operating range / quantity 20–5000 mm; selectable 1.5/3/5 m ranges Error / dispersion Manufacturer lists 1% accuracy and 10 mm resolution; the shop does not explain the percentage reference Interface I²C / Gravity, SKU SEN0304; not the analogue version Specifications, availability & sources Rate / points Up to 50 Hz; shorter range reduces cycle time and sensitivity Weight / housing Not verified · Exposed PCB/transducers; no verified IP rating Supply / power 3.3–5.5 V DC, 20 mA Field of view / coverage Manufacturer beam diagram; account for target size and range settings Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked SEN0304 listed at USD 12.90 in the manufacturer shop; confirm stock Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [DFRobot: URM09 I²C specifications and price ↗](https://www.dfrobot.com/product-1832.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1832.html) UltrasonicCompare: DFRobot URM14 DFRobot ### URM14 Close distance / glass test 99 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. Initial test-rig candidate for local distance, particularly on glass. Assessment at 50 cm sensor distance: 450–550 mm lies within the nominal range. First resolve conflicting manufacturer specifications (shop ±0.3%, wiki ±0.1%) and test rotor airflow, spray water and angled glass on a rig. Surface distance #### Strengths - Ultrasound can also detect transparent surfaces. - Affordable for a fixed close-range test setup, with an IP65 housing. #### Weaknesses - Smooth glass can reflect the echo away at oblique incidence. - Shop and wiki disagree on the error specification; rotor airflow and wetting are untested. IntegrationMedium Allow for an RS485 transceiver or bridge, Modbus decoding and a rangefinder data path checked against the specific firmware version. Operating range / quantity 0.10–1.50 m Error / dispersion Shop: ±0.3% measurement error; wiki instead states ±0.1%. Clarification required. Interface RS485 / Modbus RTU Specifications, availability & sources Rate / points 30 Hz Weight / housing Not verified · IP65 Supply / power 7–15 V DC, up to 350 mA instantaneous current Field of view / coverage 12° ±2° (-6dB) Reflections from smooth glass depend on angle. Rotor airflow, droplets on the membrane and spray mist require testing. Availability when checked Publicly priced; confirm stock. Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [DFRobot shop: technical data and price ↗](https://www.dfrobot.com/product-2173.html) - [DFRobot wiki with differing introduction ↗](https://wiki.dfrobot.com/sen0358/) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-2173.html) Industrial laser distance sensorCompare: Leuze electronic ODSL 30/D232.01-30M-S12 (50147335) Leuze electronic ### ODSL 30/D232.01-30M-S12 (50147335) Precision reference; weight unfavourable for small drones 3,118.08 USDUS shop price; tax and shipping not stated Well documented measurement reference; calculate payload and size before flight use. Assessment at 50 cm sensor distance: 450–550 mm lies within range. Static ±2/±5 mm looks promising; high weight, referencing, motion lag and glass backscatter may nevertheless rule out drone use. Surface distance #### Strengths - Millimetre error specification and broad measurement range. - Integrated polarisation filter for glossy materials. #### Weaknesses - 650 g and 79 × 69 × 150 mm do not make a lightweight drone solution. - Allow for 30–100 ms response and temperature drift up to 0.5 mm/K. IntegrationHigh 24 V DC/DC supply, RS232 level converter and protocol adapter to the flight controller; RS232 is not TTL UART. Operating range / quantity 0.20–30 m at 6–90% diffuse reflectance; up to 65 m at 90% Error / dispersion ±2 mm at 90% or ±5 mm at 6% diffuse reflectance; observe referencing requirements. Resolution 1 mm; repeatability 2 mm (3σ). Interface RS232; two push-pull switching outputs; M12, 8-pin Specifications, availability & sources Rate / points 30–100 ms response time Weight / housing 650 g net · IP67; Class 2 laser Supply / power 10–30 V DC Field of view / coverage Collimated spot beam; 6 mm at 10 m Reflectance specifications apply to diffuse reference targets, not universally to transparent wet windows; the filter alone does not guarantee glass distance measurements. Availability when checked Exact variant listed by Leuze and an authorised retailer; request lead time Purchase / region Leuze EU/international; public price from USA - [Leuze: ODSL 30/D232.01-30M-S12 technical data ↗](https://www.leuze.com/en-uk/odsl-30-d232.01-30m-s12/50147335) - [Excel Automation: exact variant and public price ↗](https://excelautomationinc.com/product/odsl-30-d232-01-30m-s12/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://excelautomationinc.com/product/odsl-30-d232-01-30m-s12/) Single-point LiDARCompare: LightWare LW20/HA LightWare ### LW20/HA Outdoor optical cross-check 389 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. Enclosed optical test candidate; US$30 more than the SF20/HA. Assessment at 50 cm sensor distance: 50 cm is within range; the HA accuracy below 1 cm applies at 20–30 °C. Temperature dependence, glass and sensor geometry can substantially reduce headroom. Surface distanceHeight above ground #### Strengths - IP67 housing at only 19.5 g. - Precise optical cross-check candidate within the specified temperature window. #### Weaknesses - Observe the temperature bands, with errors up to less than 10 cm at extreme temperatures. - Housing protection prevents neither measuring through glass nor interference from mist echoes. IntegrationLow Use UART/I²C with a verified autopilot driver; measurement direction, mounting and sensor validity must match the controller. Operating range / quantity 0.20–100 m Error / dispersion <1 cm at 20–30 °C; <2 cm at 15–20 / 30–35 °C; <5 cm at 0–15 / 35–50 °C; <10 cm at −20–0 / 50–60 °C. Interface UART / I²C Specifications, availability & sources Rate / points 48–1,250 Hz Weight / housing 19.5 g · IP67 Supply / power No reliable figure in the reviewed excerpt; check manufacturer datasheet before design Field of view / coverage Single beam; first and last returns can be evaluated Error limit remains temperature dependent. IP67 establishes housing protection, not measurement quality on glass or in spray mist. Availability when checked NW Blue lists LW20/HA at USD389 as In stock and includes Germany in its country menu; confirm final price and shipping. Manufacturer shop itself: backorder. Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [LightWare manufacturer shop: specifications and price ↗](https://lightwarelidar.com/shop/lw20-ha/) - [Price source checked on 23 October 2026 ↗](https://nwblue.com/products/lw20-ha-100-m) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://nwblue.com/products/lw20-ha-100-m) Single-point LiDARCompare: LightWare SF20/HA LightWare ### SF20/HA Precise optical cross-check 359 USDExcluding taxes; available on backorder Lightweight optical cross-check sensor for suitable, non-transparent surfaces. Assessment at 50 cm sensor distance: 50 cm can be measured. Below 1 cm is only specified at 20–30 °C; at 0–15 °C it is already below 5 cm. Test glass, spray water and dynamic flight errors separately. Surface distanceHeight above ground #### Strengths - Very light at 7.5 g, with a high available measurement rate. - Manufacturer specifies error below 1 cm under suitable conditions at 20–30 °C. #### Weaknesses - Substantially wider error limits outside the favourable temperature window. - No IP rating in the reviewed product excerpt and no demonstrated operation on wet glass. IntegrationLow UART/I²C and manufacturer autopilot guidance simplify the data path. Protection, mounting orientation and parameter settings remain project specific. Operating range / quantity 0.20–100 m Error / dispersion <1 cm at 20–30 °C; <2 cm at 15–20 / 30–35 °C; <5 cm at 0–15 / 35–50 °C; <10 cm at −20–0 / 50–60 °C. Interface USB / UART / I²C Specifications, availability & sources Rate / points 48–1,250 Hz Weight / housing 7.5 g · No IP rating stated Supply / power No reliable figure in the reviewed excerpt; check manufacturer datasheet before design Field of view / coverage Single beam; first and last returns can be evaluated Temperature-dependent error limit. No evidence of reliable measurements on wet glass. Availability when checked Available on manufacturer backorder Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [LightWare manufacturer shop: specifications and price ↗](https://lightwarelidar.com/shop/sf20-ha/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://lightwarelidar.com/shop/sf20-ha/) UltrasonicCompare: MaxBotix MB1242 I2CXL-MaxSonar-EZ4 MaxBotix ### MB1242 I2CXL-MaxSonar-EZ4 Compact I²C ultrasound for large targets 34.95 USDManufacturer price; check tax, shipping and import charges separately Software/laboratory comparison for larger flat targets; water protection requires separate development. Assessment at 50 cm sensor distance: 450–550 mm is above the 200 mm minimum distance. A 1 cm output step does not establish ±1 cm accuracy; measure temperature and target-size effects on the bench. Surface distance #### Strengths - Affordable digital I²C module with modest bench integration work. - The EZ4 variant targets large objects and greater noise tolerance within its family. #### Weaknesses - No speed-of-sound temperature correction or target-size compensation. - Exposed construction is vulnerable to water; maximum update rate depends on the environment. IntegrationMedium Use I²C pull-ups compatible with 3.3 V, short wiring and an appropriate firmware driver/bridge. The sensor may NACK while measuring; handle measurement age correctly. Operating range / quantity 200–7650 mm; reports the nearest detectable target Error / dispersion 1 cm resolution; no verified general absolute error bound. Target size can bias the result. Interface I²C; configurable address Specifications, availability & sources Rate / points Up to 40 Hz under suitable conditions; above 10 Hz reflective environments may produce old-echo readings Weight / housing Not verified · Exposed PCB and transducer; no verified IP rating Supply / power 3–5.5 V DC; measure actual current in the assembly Field of view / coverage Narrowest beam in the I2CXL-EZ family; still depends on target size and distance Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Listed at USD 34.95 in the manufacturer shop; confirm stock Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [MaxBotix: MB1242 specifications and price ↗](https://maxbotix.com/products/mb1242) - [MaxBotix: I2CXL-EZ datasheet with temperature and target-size limitations ↗](https://cdn.shopify.com/s/files/1/0550/8091/0899/files/I2CXL-MAXSONAR-EZ_DATASHEET-PD11848.pdf?v=1699290657) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://maxbotix.com/products/mb1242) UltrasonicCompare: MaxBotix MB7066-100 XL-MaxSonar-WRL1 MaxBotix ### MB7066-100 XL-MaxSonar-WRL1 Weather-resistant 10 m ultrasonic comparison 99.95 USDManufacturer starting price for the standard housing; options, tax and shipping extra Affordable outdoor comparison with ample range; account for absent temperature compensation and its signal format. Assessment at 50 cm sensor distance: 450–550 mm is outside the MB7066-100 minimum distance of 200 mm. The family datasheet does not establish general ±1 cm absolute accuracy; do not transfer another MaxBotix model’s 500 mm limit to this SKU. Surface distance #### Strengths - Full-horn enclosure with a nominal range well beyond 50 cm. - Several simultaneous outputs and low average current consumption. #### Weaknesses - No speed-of-sound compensation for temperature changes. - Serial output is not ordinary TTL UART; analogue output has 2 cm steps. IntegrationMedium Order the exact full-horn -100 configuration. A UART input needs appropriate inversion and 3.3 V level conversion; pulse-width or ADC input is an alternative. Verify temperature correction and the rangefinder bridge. Operating range / quantity 200–10,680 mm in the full-horn housing; target-dependent Error / dispersion 1 cm serial increments, 2 cm analogue steps; no verified absolute error bound for flight Interface Pulse width, analogue voltage, inverted RS232 data format at 0–Vcc levels Specifications, availability & sources Rate / points 10 Hz Weight / housing Nominal 50 g in the full-horn housing · IP67 with correct installation; protect the exposed rear/pin connections separately Supply / power 3–5.5 V DC; about 3.4 mA average and 100 mA peak at 5 V Field of view / coverage Target-dependent acoustic beam according to manufacturer diagrams Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked MB7066 listed from USD 99.95 in the manufacturer shop; confirm the -100 option and stock Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [MaxBotix: MB7066 housing options and price ↗](https://maxbotix.com/products/mb7066) - [MaxBotix: XL-WR datasheet, minimum distance and signal format ↗](https://cdn.shopify.com/s/files/1/0550/8091/0899/files/11838.pdf?v=1726579738) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://maxbotix.com/products/mb7066) Weather-resistant ultrasonicCompare: MaxBotix MB7389-100 HRXL-MaxSonar-WRMT MaxBotix ### MB7389-100 HRXL-MaxSonar-WRMT Affordable supplementary sensor; limited suitability for 45–55 cm 109.95 USDManufacturer starting price for standard housing; tax, shipping/import and options extra Simple, low-power supplementary measurement; preferably use a sensor distance above 50 cm. Assessment at 50 cm sensor distance: The 450–550 mm target band goes below the manufacturer recommendation of 500 mm for highest reliability. Therefore not a first choice as the sole control sensor; consider moving the sensor so its measurement path always remains >500 mm. Surface distance #### Strengths - Public manufacturer price and low supply voltage. - Serial millimetre values simplify a custom measurement bridge. #### Weaknesses - Manufacturer advises against distances below 50 cm when 100% measurement reliability is required. - Strongest-echo reporting and internal filters can distort edges/secondary targets or motion. IntegrationMedium Filtered 5 V supply, suitable TTL levels and firmware bridge; do not connect a 5 V output to a 3.3 V input without checking compatibility. Calibrate sensor origin relative to nozzle. Operating range / quantity 300–5000 mm; below 300 mm typically outputs 300 mm Error / dispersion Typically ±1% or better according to series datasheet; serial resolution 1 mm, analogue resolution 5 mm. Housing and near field affect errors. Interface TTL serial 9600 baud, pulse width and analogue voltage Specifications, availability & sources Rate / points 6.66 Hz with stability filter Weight / housing About 50 g in Full Horn housing; confirm selected housing option · IP67 when mounted correctly; separately seal pins/cable connection Supply / power 2.7–5.5 V DC; manufacturer shop states 2.9 mA average; some performance loss below 5 V Field of view / coverage Target-dependent acoustic beam; reports strongest echo, not necessarily the nearest object Manufacturer describes additional near-field effects at 30–50 cm. Glass angle, spray water and rotor airflow are not universally covered. Availability when checked Manufacturer lists US$109.95 and ordering options; confirm exact -100 configuration and lead time Purchase / region US manufacturer with international distributors; EU import costs unclear - [MaxBotix: MB7389 variants, price and technical overview ↗](https://maxbotix.com/products/mb7389) - [MaxBotix: current HRXL datasheet with near-range limits ↗](https://cdn.shopify.com/s/files/1/0550/8091/0899/files/11500.pdf?v=1776367207) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://maxbotix.com/products/mb7389) UltrasonicCompare: Pepperl+Fuchs UC1000-18GS-IUEP-IO-V15 (304928-100003) Pepperl+Fuchs ### UC1000-18GS-IUEP-IO-V15 (304928-100003) Compact M18 ultrasound with IO-Link diagnostics 243.20 EURDistributor price excluding VAT and shipping Compact candidate with digital diagnostics and a shorter documented response. Assessment at 50 cm sensor distance: 450–550 mm is within nominal range. ±1% of 1 m means ±10 mm characteristic deviation; also account for the separate temperature influence. The datasheet specifies two minutes of warm-up with enhanced compensation. Surface distance #### Strengths - 45 g, short blind zone and adjustable acoustic beam. - IO-Link measurements and diagnostics alongside the analogue output. #### Weaknesses - Accuracy percentages refer to the 1 m full-scale value, not to 500 mm. - Requires an IO-Link master/bridge and a regulated higher-voltage supply. IntegrationHigh Use a 24 V DC/DC converter and IO-Link master with a companion bridge, or a safely scaled ADC. Do not equate a 3 ms IO-Link cycle with the acoustic response delay. Operating range / quantity 70–1000 mm; adjustable evaluation range 90–1000 mm Error / dispersion Characteristic deviation ≤±1% of full scale; repeatability ≤±0.1% of full scale; temperature influence ≤±0.75% of full scale after warm-up Interface IO-Link 1.1 COM2, 4–20 mA/0–10 V, push-pull, synchronization; 5-pin M12 Specifications, availability & sources Rate / points Minimum 28 ms response delay, factory setting 56 ms; sensor cycle ≥14 ms Weight / housing 45 g · IP67 Supply / power 10–30 V DC, ≤1 W, ≤60 mA no-load current Field of view / coverage Adjustable beam width; target dependence remains Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Manufacturer datasheet dated May 2026 and public single-unit price; delivery time on request Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [Pepperl+Fuchs: 304928-100003 datasheet with full-scale references ↗](https://files.pepperl-fuchs.com/webcat/navi/productInfo/pds/304928-100003_eng.pdf) - [Lieske: exact variant, single-unit price and tax status ↗](https://industry-electronics.de/pepperlfuchs/304928-100003-uc1000-18gs-iuep-io-v15-ultraschallsensor-lieske_833777.htm) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://industry-electronics.de/pepperlfuchs/304928-100003-uc1000-18gs-iuep-io-v15-ultraschallsensor-lieske_833777.htm) Industrial ultrasonicCompare: Pepperl+Fuchs UC2000-30GM-IUEP-IO-V15-Y (70137787) Pepperl+Fuchs ### UC2000-30GM-IUEP-IO-V15-Y (70137787) Test direct close-range distance, including transparent surfaces Price on requestNo reliable public price verified; quotation required Lightweight industrial ultrasonic sensor with sufficient close range; interesting URM14 alternative. Assessment at 50 cm sensor distance: 450–550 mm is well outside the blind zone. At 2 m full scale, 0.2% calculates to 4 mm, but 1.5% temperature influence is 30 mm; this is not a universal ±4 mm total accuracy. Surface distance #### Strengths - 72 g and a 90 mm blind zone geometrically suit 50 cm. - Selectable acoustic beam and synchronisation support multiple sensor heads. #### Weaknesses - Percentages refer to full scale, not automatically to 500 mm. - Analogue smoothing and echo behaviour limit usable dynamics. IntegrationHigh Regulated 24 V plus IO-Link master or current/voltage ADC and a custom rangefinder bridge; output cycle is not equivalent to measurement response. Operating range / quantity 90–2000 mm; adjustable evaluation range from 120 mm Error / dispersion Characteristic deviation ≤0.2% of full scale; repeatability ≤0.1% of full scale; temperature influence up to 1.5% of full scale. Resolution ≥0.35 mm. Interface IO-Link 1.0, 4–20 mA or 0–10 V, synchronisation; M12, 5-pin Specifications, availability & sources Rate / points Response delay min. 65 ms, factory setting 125 ms; IO-Link cycle min. 33.6 ms Weight / housing 72 g · IP67 Supply / power 10–30 V DC, ≤1 W; voltage output from 15 V Field of view / coverage Adjustable acoustic beam; target-size-dependent diagram rather than a fixed FOV angle Smooth glass at oblique incidence can cause echo loss; test rotor airflow, temperature gradients, mist and a wet membrane. Availability when checked Current manufacturer datasheet and specific order number available; request price and lead time from sales Purchase / region EU and international through Pepperl+Fuchs sales - [Pepperl+Fuchs: datasheet 70137787 with error references and weight ↗](https://files.pepperl-fuchs.com/webcat/navi/productInfo/pds/70137787_eng.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://files.pepperl-fuchs.com/webcat/navi/productInfo/pds/70137787_eng.pdf) Single-point ToF development boardCompare: Pololu / STMicroelectronics VL53L1X Carrier · Pololu 3415 Pololu / STMicroelectronics ### VL53L1X Carrier · Pololu 3415 Affordable 50 cm laboratory test with adjustable field of view 22.95 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Affordable rig for matt surfaces and shading; not a ready-to-use outdoor sensor. Assessment at 50 cm sensor distance: 50 cm is generally within range; typical ±20–25 mm mathematically leaves some headroom. A laboratory candidate only: allowance for offset, angle, latency, flight motion and outdoor conditions remains unproven. Surface distanceHeight above ground #### Strengths - Very affordable, below 1 g, with regulator/level shifting for fast laboratory integration. - ROI enables controlled experiments with a narrower field of view. #### Weaknesses - Wide field of view can simultaneously detect frames, the nozzle or multiple facade sections. - Sunlight, cover windows and wetness reduce range or produce unwanted returns; custom sealing required. IntegrationHigh Develop custom firmware, timing budget, offset/crosstalk calibration and cover window; do not assume a ready-made flight-controller driver. Operating range / quantity 0.04–4 m in favourable dark conditions; Short mode typically up to 1.3 m with higher ambient-light immunity. Error / dispersion ST datasheet: typically ±20 mm in darkness and ±25 mm in the defined ambient-light tests; depends on mode, timing budget and target. Interface I²C; XSHUT and interrupt Specifications, availability & sources Rate / points Pololu driver: up to 50 Hz Short; 30 Hz Medium/Long. Do not assume the 100 ms datasheet accuracy at the maximum rate. Weight / housing 0.5 g without headers · Open board without IP rating Supply / power Carrier: 2.6–5.5 V; 20 mA typical, up to 40 mA peak Field of view / coverage 27° full; ROI can reduce this to about 15° Typical laboratory table values are not a worst-case guarantee for glass/sunlight/spray. The device does not produce a simultaneously dense depth map. Availability when checked Pololu: active/preferred; US$22.95; backorders allowed, immediate stock unconfirmed. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [Pololu: carrier 3415, price and board parameters ↗](https://www.pololu.com/product/3415) - [STMicroelectronics: VL53L1X datasheet, provided by Pololu ↗](https://www.pololu.com/file/0J1506/vl53l1x.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.pololu.com/product/3415) Single-point ToF development boardCompare: Pololu / STMicroelectronics VL53L4CD Carrier · Pololu 3692 Pololu / STMicroelectronics ### VL53L4CD Carrier · Pololu 3692 Short-range laboratory test at 50 cm 13.95 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Very affordable close-range experiment on dry, matt reference surfaces; test outdoor capability first. Assessment at 50 cm sensor distance: At 50 cm on a white target, calculated typical error is ±15 mm indoors and about ±40 mm outdoors overcast. Grey 17% targets at 50 cm already exceed the typical range at 90% valid readings. Nozzle control within ±5 cm remains unproven. Surface distance #### Strengths - Short minimum distance and low weight facilitate experiments near the nozzle. - Regulator and level shifting are already included on the affordable board. #### Weaknesses - At 50 cm, ±8% already means about ±4 cm, leaving little allowance for flight and mounting errors. - A dark target in outdoor light can fall outside the valid range before 50 cm; no sealing. IntegrationHigh Design a custom enclosure, temperature/offset calibration, measurement-status handling and timing budget; not a ready-made drone distance controller. Operating range / quantity 1 mm–1.2 m under favourable conditions; valid range can fall below 50 cm in overcast outdoor light with a dark target. Error / dispersion ST at 33 ms and >200 mm: indoors typically ±3% (white) / ±4% (grey); outdoors overcast typically ±8%. At least 90% of readings within the stated limits. Interface I²C; XSHUT and interrupt Specifications, availability & sources Rate / points Up to 100 Hz; 33 ms accuracy figures do not automatically apply at 100 Hz. Weight / housing 0.5 g without headers · Open board without IP rating Supply / power Carrier: 2.6–5.5 V; 25 mA typical, up to 40 mA peak Field of view / coverage 18° diagonal 1 mm resolution does not mean 1 mm accuracy. At 33 ms, ST states a typical range of 450 mm indoors for grey 17% targets at 90% valid readings, and only 400 mm outdoors overcast; 50 cm is not promised there. Availability when checked Pololu: active/preferred; US$13.95; backorders allowed, immediate stock unconfirmed. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [Pololu: carrier 3692, price and board parameters ↗](https://www.pololu.com/product/3692) - [STMicroelectronics: VL53L4CD datasheet Rev. 8, provided by Pololu ↗](https://www.pololu.com/file/0J2060/vl53l4cd.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.pololu.com/product/3692) Industrial laser distance sensorCompare: SICK DT35-B15251 (1057652) SICK ### DT35-B15251 (1057652) Direct facade distance; lightweight industrial laser 701.22 USDUS shop price includes stated tariffs; tax and shipping depend on destination Interesting rig candidate for 50 cm on diffusely reflecting surfaces. Assessment at 50 cm sensor distance: 450–550 mm lies within range. Typically ±10 mm does not guarantee ±5 cm nozzle guidance: incidence angle, mounting offset, flight motion and latency add further errors. Surface distance #### Strengths - 65 g is favourable for an industrial sensor. - Typical measurement error is well below the 50 mm target band. #### Weaknesses - 40 klx ambient-light immunity and glass reflections limit outdoor applications. - Allow 20 minutes of warm-up for specified conditions. IntegrationHigh Regulated 24 V supply plus analogue ADC or IO-Link master and rangefinder bridge; not a direct TTL UART device. Operating range / quantity 0.05–12 m at 90% remission and Speed=Slow; up to 3.1 m at 6% Error / dispersion Typically ±10 mm; resolution 0.1 mm. Repeatability is separate and depends on speed. Interface IO-Link 1.1, 4–20 mA or 0–10 V; M12, 5-pin Specifications, availability & sources Rate / points 2.5–96.5 ms response time; output every 1–32 ms depending on setting Weight / housing 65 g, without bracket/cable · IP65/IP67; Class 2 laser Supply / power 12–30 V DC, ≤1.7 W; IO-Link requires >18 V Field of view / coverage Single measurement spot; typically 15 × 15 mm at 2 m Test glass, wet/glossy surfaces and spray mist in the measurement path; IP67 does not guarantee a valid reading with wet optics. Availability when checked Newark listed 10 units; confirm stock and EU delivery before ordering Purchase / region SICK international; price source USA - [SICK: DT35-B15251 datasheet, 1057652 ↗](https://www.sick.com/media/pdf/3/53/353/dataSheet_DT35-B15251_1057652_en.pdf) - [Newark: DT35-B15251 unit price and stock ↗](https://www.newark.com/sick/dt35-b15251/photo-sensor-12m-laser-npn-pnp/dp/54AK5716) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.newark.com/sick/dt35-b15251/photo-sensor-12m-laser-npn-pnp/dp/54AK5716) UltrasonicCompare: SICK UM18-218126111 (6048404) SICK ### UM18-218126111 (6048404) Lightweight M18 industrial sensor for 50 cm 326.23 EURRS Netherlands: excluding VAT and shipping A useful lightweight industrial comparison with URM14 and pico+100/I. Assessment at 50 cm sensor distance: 450–550 mm is within operating range. ±1% calculates to ±5 mm at 500 mm under manufacturer conditions; this does not prove dynamic ±5 cm performance in wind and spray. Surface distance #### Strengths - 25 g with accuracy explicitly referenced to measured distance. - 120 mm blind zone and synchronization for multiple heads. #### Weaknesses - Requires a 24 V supply and a 4–20 mA measurement input. - The 20 ms output interval is faster than the actual 80 ms response. IntegrationHigh Provide regulated 24 V, a current receiver/ADC with a 3.3 V-safe output, and a rangefinder bridge. Measure filtering, timestamps and multiplex rate for the complete control loop. Operating range / quantity 120–1000 mm operating range; 1300 mm limiting range Error / dispersion ±1% of current measured distance, ±0.15% repeatability; resolution ≥0.069 mm Interface 4–20 mA, 5-pin M12; synchronization/multiplexing Specifications, availability & sources Rate / points 80 ms response, 20 ms output interval; smoothing may increase response time Weight / housing 25 g · IP65/IP67 Supply / power 10–30 V DC, ≤1.2 W without load Field of view / coverage Target-dependent acoustic beam according to manufacturer diagram Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Exact variant appears in the manufacturer catalogue and RS single-unit listing; confirm delivery Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [SICK: current UM18-218126111 datasheet ↗](https://www.sick.com/media/pdf/7/67/767/dataSheet_UM18-218126111_6048404_en.pdf) - [RS Netherlands: single-unit price for the exact variant ↗](https://nl.rs-online.com/web/p/proximity-sensors/2125525) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://nl.rs-online.com/web/p/proximity-sensors/2125525) Industrial ultrasonicCompare: SICK UM30-213113 (6036918) SICK ### UM30-213113 (6036918) Direct distance measurement with a clear error reference 1,059.65 USDUS shop price; tax and shipping not stated Industrial alternative with documented measurement accuracy; relatively expensive. Assessment at 50 cm sensor distance: At 500 mm, ±1% calculates to ±5 mm under manufacturer conditions. 450–550 mm lies within the operating range; dynamic nozzle guidance within ±5 cm requires separate verification. Surface distance #### Strengths - ±1% explicitly refers to the current distance. - Configurable synchronisation/multiplexing and measurement filters. #### Weaknesses - 150 g and high unit price make multiple measurement directions harder to implement. - A 23 ms output interval does not mean a 23 ms response to a distance change. IntegrationHigh Allow for regulated 24 V and an analogue ADC with MAVLink/rangefinder bridge; account for filters and timestamps in flight testing. Operating range / quantity 200–1300 mm operating range; limiting range 2000 mm Error / dispersion ±1% of the current measured value; repeatability ±0.15%; resolution ≥0.18 mm Interface 4–20 mA or 0–10 V, synchronisation/multiplexing; M12, 5-pin Specifications, availability & sources Rate / points 92 ms response; 23 ms output interval. Additional smoothing may extend response. Weight / housing 150 g · IP65/IP67 Supply / power 9–30 V DC, ≤2.4 W; 0–10 V output at 15–30 V Field of view / coverage Adjustable acoustic beam; target detection according to manufacturer diagram Temperature compensation does not establish robustness against rotor airflow or a wet membrane; echoes can be lost on smooth angled glass. Availability when checked RS USA listed 16 units with a public price; obtain EU quotation separately Purchase / region SICK international; price source USA - [SICK: UM30-213113 datasheet with percentage references ↗](https://www.sick.com/media/pdf/7/57/357/dataSheet_UM30-213113_6036918_en.pdf) - [RS: UM30-213113 unit price and stock ↗](https://us.rs-online.com/product/sick/um30-213113/70875246/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://us.rs-online.com/product/sick/um30-213113/70875246/) Industrial laser distance sensorCompare: ifm electronic O1D100 / O1DLF3KG/IO-LINK ifm electronic ### O1D100 / O1DLF3KG/IO-LINK Direct distance with documented light/colour errors 570.90 USDUS list price; taxes, shipping and possible duties extra Well documented industrial comparison for diffuse facades, with higher weight. Assessment at 50 cm sensor distance: 50 cm falls within the error specifications for 200–1000 mm. ±15–26.5 mm leaves static allowance relative to the ±50 mm goal; total flight error and measurement dropouts are not established by this. Surface distance #### Strengths - Error tables account for 50 Hz and strong ambient light. - Analogue output and IO-Link facilitate reproducible rig measurements. #### Weaknesses - About 244 g plus cabling and converters adds aircraft load. - Reflective/wet glass is not covered by error tables based on paper targets. IntegrationHigh 24 V DC/DC, 4–20 mA input or IO-Link master and bridge required; transfer readings to the flight controller with timestamps and validity checks. Operating range / quantity 0.20–10 m on white reference target; black up to 4 m Error / dispersion At 0.2–1 m/50 Hz: ±15 mm white or ±18 mm grey below 40 klx; ±26.5 mm at 40–100 klx. This is not repeatability. Interface IO-Link 1.1, 4–20 mA/0–10 V, PNP; M12 Specifications, availability & sources Rate / points 1–50 Hz; accuracy depends on sampling rate and light Weight / housing 243.5 g · IP67; Class 2 laser Supply / power 18–30 V DC; <150 mA Field of view / coverage Single measurement spot; max. 25 mm at 10 m Reference data applies under constant ambient conditions and after warm-up; separately test raindrops, oblique views and spray mist. Availability when checked DigiKey listed 27 units; confirm internationally deliverable variant and lead time Purchase / region ifm international; price source USA - [ifm: O1D100 technical data ↗](https://www.ifm.com/gb/en/product/O1D100) - [ifm: datasheet with 50 Hz error tables ↗](https://media.ifm.com/dam/5d768386-a477-443e-969c-ce7e978f0937/Original/O1D100-06_EN-GB.pdf) - [DigiKey: O1D100 unit price and stock ↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/O1D100/12144829) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/O1D100/12144829) UltrasonicCompare: microsonic lpc+100/CFI (36310) microsonic ### lpc+100/CFI (36310) Lightweight IO-Link ultrasound with analogue option 301.30 EURFrench distributor price excluding VAT and shipping Compact industrial comparison with a digital distance value rather than only a switching signal. Assessment at 50 cm sensor distance: 450–550 mm falls within operating range and above the 120 mm blind zone. Temperature-compensated accuracy does not establish performance in rotor airflow and spray. Surface distance #### Strengths - 35 g with an IO-Link distance value and scaled process data. - Analogue output, synchronization and multiplexing are also available. #### Weaknesses - An IO-Link master or analogue adapter adds system weight and cost. - A 20 ms data cycle does not mean a 20 ms distance response. IntegrationHigh Use regulated 24 V plus an IO-Link master and companion bridge, or a 4–20 mA ADC. Decode process-data scaling and error states according to the IODD; never connect 24 V directly to the Cube. Operating range / quantity 120–1000 mm operating range; 1300 mm limiting range Error / dispersion ±1% with internal temperature compensation; ±0.15% repeatability Interface IO-Link COM2 distance data, 4–20 mA, push-pull; synchronization/multiplexing; 5-pin M12 Specifications, availability & sources Rate / points 80 ms response delay; minimum IO-Link cycle 20 ms Weight / housing 35 g · IP67 Supply / power 10–30 V DC, ≤50 mA without load Field of view / coverage Target-dependent manufacturer acoustic beam diagram Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Current manufacturer page and public single-unit price at GSD Automatisme; confirm delivery Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [microsonic: lpc+100/CFI specifications and IO-Link process data ↗](https://www.microsonic.de/en/sensordetail/lpc%2B100-CFI) - [GSD Automatisme: exact CFI variant and net single-unit price ↗](https://www.gsd-automatisme.com/produit/80179/capteurs-a-ultrasons/cylindriques/microsonic/lpc100/cfi) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.gsd-automatisme.com/produit/80179/capteurs-a-ultrasons/cylindriques/microsonic/lpc100/cfi) Industrial ultrasonicCompare: microsonic mic+130/IU/TC (22 320) microsonic ### mic+130/IU/TC (22 320) Precise ultrasonic comparison on a facade test rig 284.70 EURNet, plus VAT and shipping Useful industrial candidate with range allowance below and above 50 cm. Assessment at 50 cm sensor distance: 450–550 mm lies within the operating range and outside the 200 mm blind zone. Distinguish ±1% from resolution and repeatability; no proof of ±5 cm in flight with the nozzle running. Surface distance #### Strengths - Actual analogue distance reading with temperature compensation. - Display and teach-in simplify test-rig setup. #### Weaknesses - 150 g plus adapter is substantially heavier than small drone rangefinders. - 92 ms response calls for slow, validated distance control. IntegrationHigh Regulated 24 V rail, current/voltage input and bridge to the flight controller; synchronise multiple sensors rather than transmitting without coordination. Operating range / quantity 200–1300 mm operating range; limiting range 2000 mm Error / dispersion ±1% with internal temperature compensation; repeatability ±0.15%; resolution 0.18–0.57 mm depending on analogue window Interface 4–20 mA/0–10 V; synchronisation/multiplexing; M12, 5-pin Specifications, availability & sources Rate / points 92 ms response delay Weight / housing 150 g · IP67 Supply / power 9–30 V DC, ≤80 mA without load; 0–10 V from 15 V Field of view / coverage Acoustic beam according to manufacturer diagram; no universal fixed angle Ultrasound detects transparent targets depending on their echoes; angled glass, rotor airflow and a water film on the transducer remain test conditions. Availability when checked Automation24 lists a specific price and next-working-day dispatch; confirm stock before purchase Purchase / region Germany/EU; international via microsonic sales - [microsonic: mic+130/IU/TC technical data ↗](https://www.microsonic.de/de/produktdetail/mic%2B130-IU-TC) - [Automation24: mic+130/IU/TC price and availability ↗](https://www.automation24.de/ultraschallsensor-microsonic-mic-130-iu-tc) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.automation24.de/ultraschallsensor-microsonic-mic-130-iu-tc) Compact industrial ultrasonicCompare: microsonic pico+100/I microsonic ### pico+100/I Lightweight direct close-range distance; glass test-rig candidate 234.10 EURConfirmed net price; taxes and shipping are additional. Especially interesting weight comparison with URM14; select the exact straight-head variant. Assessment at 50 cm sensor distance: 450–550 mm lies well within the operating range. ±1% remains a sensor specification under measurement conditions; verify the entire nozzle distance band with motion, spray water and signal dropouts. Surface distance #### Strengths - Only 30 g with an industrial IP67 housing. - 120 mm blind zone leaves room for a 50 cm working distance. #### Weaknesses - This I variant has an analogue output, not an IO-Link distance telegram. - Allow for 80 ms response plus bridge latency in the control loop. IntegrationHigh 24 V DC/DC and 4–20 mA receiver/ADC plus rangefinder bridge required. Retailer short description incorrectly mentions an angled head; manufacturer distinguishes the separate /WK/I variant. Operating range / quantity 120–1000 mm operating range; limiting range 1300 mm Error / dispersion ±1% with temperature compensation; repeatability ±0.15%; resolution 0.069–0.38 mm depending on analogue window Interface 4–20 mA; teach-in/synchronisation through Com input; M12, 5-pin Specifications, availability & sources Rate / points 80 ms response delay Weight / housing 30 g · IP67 Supply / power 10–30 V DC, ≤40 mA without load Field of view / coverage Target-size-dependent acoustic beam according to manufacturer diagram Test echo quality on angled glass, wetting and rotor airflow. Very fine resolution must not be presented as measurement accuracy. Availability when checked Dealer lists stock; confirm straight sensor head. Purchase / region EU/Netherlands; manufacturer Germany - [microsonic: pico+100/I technical data, 30 g, straight head ↗](https://www.microsonic.de/de/produktdetail/pico%2B100-I) - [Sensor Partners: pico+100/I price and stock ↗](https://sensorpartners.com/en/product/microsonic-pico100-i/) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://sensorpartners.com/en/product/microsonic-pico100-i/) UltrasonicCompare: microsonic zws-70/CI/QS microsonic ### zws-70/CI/QS Very lightweight industrial ultrasound for 50 cm 234.10 EURConfirmed net price; taxes and shipping are additional. A useful lightweight bench candidate when a 24 V analogue interface is already available. Assessment at 50 cm sensor distance: 450–550 mm is well beyond the 120 mm blind zone and within operating range. The ±1% sensor specification and 0.20 mm resolution do not establish guaranteed flight accuracy. Surface distance #### Strengths - Only 11 g sensor weight in an enclosed industrial housing. - 450–550 mm falls within the 700 mm operating range. #### Weaknesses - Requires 20–30 V and a separate 4–20 mA receiver. - 70 ms response; synchronization is available, but no native multiplex mode. IntegrationHigh Provide a 24 V DC/DC converter, 4–20 mA receiver and 3.3 V-safe ADC/rangefinder bridge; deliberately schedule external synchronization for multiple heads. Operating range / quantity 120–700 mm operating range; 1000 mm limiting range Error / dispersion ±1% with internal temperature compensation; ±0.15% repeatability, 0.20 mm resolution Interface 4–20 mA, 4-pin M8; external synchronization Specifications, availability & sources Rate / points 70 ms response delay Weight / housing 11 g · IP67 Supply / power 20–30 V DC, ≤30 mA without load Field of view / coverage Target-dependent acoustic beam according to manufacturer diagram Smooth angled glass may reflect echoes away. Test rotor airflow, spray, water films and changing temperatures on the actual assembly; an IP rating does not demonstrate performance during pressure washing. Availability when checked Listed at EUR 234.10 excluding VAT by Sensor Partners; confirm stock Purchase / region International manufacturer/distributor channels; confirm delivery to Germany and the final cost - [microsonic: zws-70/CI/QS specifications, 11 g and operating range ↗](https://www.microsonic.de/en/sensordetail/zws-70-CI-QS) - [Sensor Partners: exact CI/QS variant and net price ↗](https://sensorpartners.com/de/produkt/microsonic-zws-70-ci-qs/) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://sensorpartners.com/de/produkt/microsonic-zws-70-ci-qs/) Single-point ToF / LiDARCompare: Benewake TF-Luna · 8 m (DFRobot SEN0340) Benewake ### TF-Luna · 8 m (DFRobot SEN0340) Affordable single-point distance for the test rig 24.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Very affordable comparison sensor for diffuse facade surfaces; mount with protection. Assessment at 50 cm sensor distance: 50 cm lies within range, but ±6 cm at 50 cm does not meet ±5 cm nozzle distance. Sensor offset, tilt, latency and flight motion add further errors. Surface distanceHeight above ground #### Strengths - Very light and low power; enables multiple measurement directions on a rig. - UART and I²C simplify integration into a data acquisition system. #### Weaknesses - ±6 cm nominal error alone exceeds the entire ±5 cm nozzle budget. - Unsealed; separately test glass, sunlight, droplets and weak return signals. IntegrationMedium Check firmware mode, supply, measurement status and mounting offset; protective housing required. Operating range / quantity 0.20–8 m at 90% reflectance; only up to 2.5 m at 10% reflectance (indoors). Error / dispersion ±6 cm at 0.2–3 m; ±2% at 3–8 m. 1 cm resolution is not an accuracy specification. Interface UART, I²C, I/O; 3.3 V logic Specifications, availability & sources Rate / points 1–250 Hz adjustable; 100 Hz default Weight / housing <5 g · No sealed housing / no IP claim Supply / power 3.7–5.2 V; ≤0.35 W Field of view / coverage 2° theoretical A single beam does not provide a facade map. Below 20 cm readings are unreliable; dark surfaces reduce range. Availability when checked DFRobot lists US$24.90 and an order button; available quantity unverified. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [DFRobot: price and SEN0340 model ↗](https://www.dfrobot.com/product-1995.html/) - [Benewake: TF-Luna user manual ↗](https://en.benewake.com/uploadfiles/2025/04/20250430174515390.pdf) - [Benewake: current TF-Luna product data ↗](https://ai.benewake.com/en/products/tf-luna/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1995.html/) Single-point ToF / LiDARCompare: Benewake TF02-Pro · UART/I²C (DFRobot SEN0238) Benewake ### TF02-Pro · UART/I²C (DFRobot SEN0238) Enclosed budget distance and ground-height sensor 87.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Affordable comparison for height measurement and local distance on well-reflecting surfaces. Assessment at 50 cm sensor distance: 50 cm is within nominal range, but ±5 cm already consumes the entire target budget. A mounting offset increasing sensor distance does not change this near-range limitation. Surface distanceHeight above ground #### Strengths - IP65 and specified 100 klux measurement conditions suit external mounting better than open modules. - 40 m on bright targets supports distances well beyond nozzle standoff. #### Weaknesses - ±5 cm sensor error leaves no allowance for flight and mounting errors within the ±5 cm total goal. - A single point may hit frames, glass or spray mist instead of the intended facade surface. IntegrationMedium Check suitable UART/I²C driver, reading age, outlier handling and sensor-to-nozzle transformation. Operating range / quantity 0.10–40 m at 90% reflectance; 0.10–13.5 m at 10%, including 100 klux. Error / dispersion ±5 cm at 0.1–5 m; ±1% at 5–40 m; reference: white surface, 25 °C. Interface UART, I²C, I/O; 3.3 V logic Specifications, availability & sources Rate / points 1–1000 Hz; 100 Hz default Weight / housing 50 g including cable · IP65 Supply / power 5–12 V; ≤1 W Field of view / coverage 3° IP65 does not establish resistance to high-pressure jets or cleaning chemicals; no validated complete solution for cleaning flight. Availability when checked DFRobot: US$87.90; “Only 9 Left” when checked. Stock is a snapshot. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [DFRobot: SEN0238 price and stock ↗](https://www.dfrobot.com/product-1599.html/) - [Benewake: TF02-Pro datasheet A02 ↗](https://en.benewake.com/uploadfiles/2024/04/20240426135509321.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1599.html/) Single-point ToF / LiDARCompare: Benewake TF02-i · RS-485 (DFRobot SEN0504) Benewake ### TF02-i · RS-485 (DFRobot SEN0504) Direct distance with more robust fieldbus cabling 109.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Candidate for longer, interference-prone wiring between sensor head and measurement computer. Assessment at 50 cm sensor distance: At 50 cm, nominal error is already ±5 cm. Offset, angle, latency and flight motion require additional allowance; no assurance of 50 cm ±5 cm. Surface distanceHeight above ground #### Strengths - RS-485 and a wide supply-voltage range facilitate robust cabling. - Up to 40 m on bright surfaces; 100 klux condition is specified in manufacturer datasheet. #### Weaknesses - Frame rate and protection rating conflict across current sources and require clarification before purchase. - ±5 cm nominal error leaves no allowance for the complete nozzle-distance goal. IntegrationHigh RS-485 is not automatically DroneCAN. Check adapter, termination and exact protocol. Operating range / quantity 0.10–40 m at 90% reflectance; up to 13.5 m at 10%, including 100 klux. Error / dispersion Manufacturer A03: ±5 cm at 0.1–5 m; ±1% at 5–25 m. No accuracy promise for 25–40 m in that document. Interface Offered SEN0504 version: RS-485; CAN is a separately orderable version. Specifications, availability & sources Rate / points 1–100 Hz according to A03 footnote; conflicting 1000 Hz line in datasheet and retailer text. Weight / housing 60 g including cable · Retailer states IP65; manufacturer datasheet A03 gives no IP rating. Clarify supplied version in writing. Supply / power 7–30 V; ≤0.85 W at 12 V Field of view / coverage 3° theoretical Do not mix retailer specifications with the CAN version. Unambiguous current manufacturer assurances are missing for accuracy above 25 m and water protection. Availability when checked DFRobot: US$109.90; “Only 8 Left” when checked. Supplied version RS-485. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [DFRobot: SEN0504 RS-485 price and stock ↗](https://www.dfrobot.com/product-2418.html) - [Benewake: TF02-i datasheet A03 with source conflict ↗](https://en.benewake.com/uploadfiles/2024/04/20240426135616492.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-2418.html) Single-point LiDARCompare: Benewake TF03-100 Standard UART/CAN Benewake ### TF03-100 Standard UART/CAN Directional distance measurement / supplementary ground or wall measurement 229.11 USDRobotShop US price includes US tariffs according to listing; VAT/sales-tax status not explicitly stated, shipping extra. Robust single-point sensor for longer distances; datasheet accuracy does not fit a stand-alone ±5 cm close-range requirement. Assessment at 50 cm sensor distance: At 50 cm, ±10 cm applies according to A01 datasheet. Insufficient for the ±5 cm total goal without an additional, more precise measurement channel. Surface distanceHeight above ground #### Strengths - Narrow measurement cone and IP67 housing. - High configurable data rate and two standard interfaces. #### Weaknesses - Only one beam; does not fully detect lateral obstacles or surface inclination. - Optical target properties, sunlight and water require testing; no verified glass detection. IntegrationMedium Confirm hardware revision and interface in writing; evaluate signal strength, dropouts and filter delay. Operating range / quantity 0.1–100 m at 90% reflectance; 40 m at 10%; at 100 klux 80 m and 30 m respectively according to A01 Error / dispersion ±10 cm below 10 m; ±1% from 10 m; 1 cm resolution is not accuracy Interface Standard UART/CAN; RS232/RS485 are other hardware variants Specifications, availability & sources Rate / points A01 standard: 1–1,000 Hz, default 100 Hz; newer series page states 9,800 Hz, clarify revision when purchasing Weight / housing A01 UART/CAN: 89 ±3 g; newer series page states 86 ±3 g without cable · IP67; no assurance against a high-pressure cleaning jet Supply / power A01: 5–24 V DC; ≤1 W Field of view / coverage 0.5° according to A01 No assured ±5 cm at 0.5 m; repeatability and resolution must not be presented as absolute accuracy. Availability when checked RobotShop: 3 units in stock when checked; conflicting technical generation details in retailer text. Purchase / region RobotShop US offer verified; check delivery region and import costs in the relevant country shop. - [Benewake TF03-100 datasheet A01, 2024 ↗](https://en.benewake.com/uploadfiles/2024/04/20240426134134211.pdf) - [Benewake current TF03 series page / revision discrepancy ↗](https://ai.benewake.com/en/resources/technical-documents/tf03-datasheet/) - [RobotShop TF03-100 RB-Ben-12: price and stock ↗](https://www.robotshop.com/products/benewake-tf03-lidar-led-rangefinder-ip67-100-m) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.robotshop.com/products/benewake-tf03-lidar-led-rangefinder-ip67-100-m) Single-point ToF / LiDARCompare: Benewake TFS20-L · 20 m OEM module (DFRobot SEN0672) Benewake ### TFS20-L · 20 m OEM module (DFRobot SEN0672) Extremely lightweight distance module for custom integration 43.90 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Interesting for low weight and multiple measurement directions when electronics and weather protection are developed in-house. Assessment at 50 cm sensor distance: 50 cm lies within range; ±6 cm absolute accuracy exceeds the total goal. 2 cm repeatability does not replace absolute error and does not establish flight-control performance. Surface distanceHeight above ground #### Strengths - Only about 1.35 g with substantially longer range than simple close-range chips. - Officially separate ranges for sunlight and dark targets enable realistic tests. #### Weaknesses - ±6 cm nominal error is insufficient for ±5 cm nozzle standoff. - Open electronics, small 0.8 mm connector and custom 3.3 V supply increase integration effort. IntegrationHigh Design housing, cable strain relief, supply headroom and protocol integration. Operating range / quantity 0.20–20 m at 90% reflectance in darkness; 15 m at 100 klux. Dark target: 12 m indoors / 9 m at 100 klux. Error / dispersion ±6 cm at 0.2–6 m; ±1% from 6 m. Repeatability 2 cm (1σ) is separate. Interface UART or I²C; 3.3 V logic Specifications, availability & sources Rate / points 20 / 50 / 100 / 250 Hz; single-measurement mode 0 Weight / housing About 1.35 g without additional mounting · N/A; open OEM module without IP rating Supply / power 3.3 V ±9%; ≤0.43 W Field of view / coverage <2° 20 m does not apply to dark targets in full sunlight. Not a sealed flight or cleaning sensor. Availability when checked DFRobot lists US$43.90 and an order button; unit stock unverified. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [DFRobot: SEN0672 price ↗](https://www.dfrobot.com/product-2951.html) - [Benewake: TFS20-L datasheet ↗](https://en.benewake.com/uploadfiles/2025/01/20250102154203449.pdf) - [Benewake: TFS20-L user manual ↗](https://en.benewake.com/uploadfiles/2025/04/20250430175419778.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-2951.html) Single-point ToFCompare: Benewake TFmini Plus Benewake ### TFmini Plus Learning / approximate distance 49.90 USDTax status not stated Affordable for driver development and initial distance experiments. Assessment at 50 cm sensor distance: 50 cm lies within range, but ±5 cm sensor error consumes the entire permissible nozzle deviation. No allowance remains when used as the sole control sensor for this goal. Surface distanceHeight above ground #### Strengths - Low weight and simple serial measurement path. - Affordable entry into driver and telemetry testing. #### Weaknesses - ±5 cm sensor error leaves no allowance for ±5 cm total deviation. - A second optical data path does not automatically resolve glass, reflections and mist. IntegrationLow Check support, interface and parameters of the specific ArduPilot/PX4 firmware; connecting the sensor alone does not create distance control. Operating range / quantity 0.10–12 m; dark targets up to 4 m Error / dispersion ±5 cm at 0.1–5 m; ±1% at 5–12 m according to A07 datasheet. Interface UART / I²C, depending on variant Specifications, availability & sources Rate / points UART up to 1,000 Hz Weight / housing 12 ±1 g · IP65 Supply / power 5 V ±0.5 V; 550 mW Field of view / coverage 3.6° nominal ±5 cm sensor error already consumes the desired total tolerance. Optical errors on glass and mist remain possible. Availability when checked Product listed with public price; check actual stock before ordering Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Benewake TFmini Plus A07 manufacturer datasheet ↗](https://en.benewake.com/uploadfiles/2024/04/20240426135842086.pdf) - [DFRobot retailer price ↗](https://www.dfrobot.com/product-1915.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1915.html) Single-point ToF with LEDCompare: Garmin LIDAR-Lite v4 LED · 010-02022-00 Garmin ### LIDAR-Lite v4 LED · 010-02022-00 Lightweight close-range distance with documented repeatability 69.99 USDPublic USD unit price; tax status not stated. Check shipping and possible import duties separately. Rig candidate for a lightweight 50 cm channel; separate absolute calibration required. Assessment at 50 cm sensor distance: 50 cm lies within range and ±1 cm repeatability sounds promising. Absolute accuracy, sunlight, wet glass, offset and latency still lack complete verification for ±5 cm. Surface distanceHeight above ground #### Strengths - Documented 1 cm repeatability up to 2 m under clearly defined indoor conditions. - 14.6 g and direct I²C connection; manufacturer still lists the model with an order option. #### Weaknesses - Repeatability is not absolute accuracy and does not establish performance on sunny, wet facades. - No verified sealing; High Accuracy averaging and radio increase or vary latency. IntegrationMedium Prefer testing wired I²C; evaluate status: an invalid return may be output as 1 cm. Operating range / quantity 0.05–10 m; reference point according to retailer: rear of the device. Error / dispersion Repeatability in High Accuracy mode indoors on a 90% target: ±1 cm up to 2 m, ±2 cm up to 4 m, ±5 cm up to 10 m (1σ). No general absolute-error guarantee. Interface I²C or ANT; SPI only through a custom application Specifications, availability & sources Rate / points I²C >200 Hz typical; averaging in High Accuracy mode reduces output rate. Weight / housing 14.6 g · No IP rating in the verified manual Supply / power 4.75–5.25 V; 85 mA during measurement, 2 mA idle Field of view / coverage 4.77° beam divergence Not a laser scanner. Account for reference point, High Accuracy setting and invalid 1 cm outputs in processing. Availability when checked SparkFun when checked: US$69.99, “In stock”. Garmin lists model and shipping information; no EOL notice found. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [SparkFun: SEN-15776 price and stock ↗](https://www.sparkfun.com/garmin-lidar-lite-v4-led-distance-measurement-sensor.html) - [Garmin: manufacturer product and lifecycle check ↗](https://www.garmin.com/en-US/p/610275/) - [Garmin: operating manual with technical data ↗](https://static.garmin.com/pumac/LIDAR-Lite%20LED%20v4%20Instructions_EN-US.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.sparkfun.com/garmin-lidar-lite-v4-led-distance-measurement-sensor.html) Single-point ToF / LiDARCompare: LightWare LW20/C · 100 m (standard, not HA) LightWare ### LW20/C · 100 m (standard, not HA) Sealed distance and height channel 349 USDManufacturer price explicitly excludes taxes; shipping and import extra. Compact weather-protected alternative for a second measurement channel; lower close-range accuracy than LW20/HA. Assessment at 50 cm sensor distance: At 50 cm, ±5 cm at low rate consumes the entire target budget; high rate is specified at ±10 cm. Not verified for ±5 cm total nozzle standoff. Surface distanceHeight above ground #### Strengths - Complete IP67 device at just under 20 g including cable. - Servo driver enables additional steering of a single beam for simple scanning experiments. #### Weaknesses - Standard C version has substantially less error headroom than the existing HA version. - Servo scanning needs custom angle sensing, synchronisation and suitable frame rate; not a complete mapping system. IntegrationMedium Test a fixed channel first; use a servo only with separate angle/time calibration. IP67 does not establish chemical/high-pressure resistance. Operating range / quantity 0.20–100 m; 70% albedo in sunlight, 0.9 × 0.9 m target. Error / dispersion ±5 cm below 500 readings/s; ±10 cm above 500/s. Do not apply HA accuracy. Interface UART, I²C, servo driver; 3.3 V TTL, 5 V tolerant Specifications, availability & sources Rate / points 48–5000 readings/s Weight / housing 19.5 g including cable · IP67 for the complete device Supply / power 4.5–5.5 V; 85 mA typical, <200 mA at startup Field of view / coverage Beam divergence <0.5° Do not transfer the HA specification to this C version. Glass, spray mist and changing targets require live testing. Availability when checked Manufacturer: US$349 excluding taxes; “Available on backorder”. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [LightWare: LW20/C standard version, price and backorder ↗](https://lightwarelidar.com/shop/lw20-c-100-m/) - [LightWare: product guide 13.4, 24 November 2025 ↗](https://lightwarelidar.com/wp-content/uploads/2025/12/LW20-Product-Guide-V13.4_Corrected.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://lightwarelidar.com/shop/lw20-c-100-m/) Single-point ToF / LiDARCompare: LightWare SF000/B · 50 m LightWare ### SF000/B · 50 m Lightweight single-point distance and height channel 319 USDManufacturer price explicitly excludes taxes; shipping and import extra. Very light additional distance channel with a narrow beam; protect the rear during integration. Assessment at 50 cm sensor distance: 50 cm can be measured, but ±5 cm sensor error leaves no allowance. Calibrate nozzle mounting offset and time delay on the complete system. Surface distanceHeight above ground #### Strengths - Below 10 g without cable at 50 m nominal range. - First and last returns support investigation of multiple targets in the beam. #### Weaknesses - ±5 cm sensor error exhausts the nozzle budget. - Only the front is IP65; do not expose the complete module unprotected to spray. IntegrationMedium Custom sealing and heat dissipation; check UART/I²C or a separate DroneCAN adapter. Operating range / quantity 0.20–50 m; 70% albedo in sunlight, 0.9 × 0.9 m target. Error / dispersion ±5 cm according to product guide 7.3; 1 cm resolution. Interface UART and I²C, 3.3 V TTL and 5 V tolerant Specifications, availability & sources Rate / points 48–388 readings/s Weight / housing 9.81 g without cable · Complete device IP00; front only IP65 Supply / power 4.5–5.5 V; 84 mA typical (about 0.42 W at 5 V) Field of view / coverage Beam divergence <1° One beam is not a 3D map. The 50 m applies to defined target and visibility conditions. Availability when checked Manufacturer: US$319 excluding taxes; “Available on backorder”, no immediate stock promised. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [LightWare: SF000/B current shop price and backorder ↗](https://lightwarelidar.com/shop/sf000-b-50-m/) - [LightWare: product guide 7.3, 4 May 2026 ↗](https://lightwarelidar.com/wp-content/uploads/2026/05/SF000-Product-Guide-v7.3.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://lightwarelidar.com/shop/sf000-b-50-m/) Radar altimeterCompare: Ainstein US-D1 Radar Altimeter Ainstein ### US-D1 Radar Altimeter Ground distance / AGL height; no facade odometry 599 USDTax status not explicitly stated in manufacturer shop; shipping extra. Candidate for redundant ground height during takeoff/landing; not recommended for frontal glass-distance protection. Assessment at 50 cm sensor distance: 50 cm is the lower measurement limit. The 450–550 mm target band intersects the blind zone; precision is also 6 cm below 1 m. Classify as an altimeter, not a precise facade-control sensor. Height above ground #### Strengths - Radar complements optical height measurement in difficult light or dust. - ArduPilot and PX4 integration are documented. #### Weaknesses - Ainstein explicitly does not recommend US-D1 as an obstacle-avoidance sensor. - Large objects and multipath echoes within the wide field of view can distort height readings. IntegrationMedium Point downward; evaluate SNR and invalid values, check installation and firmware variant. Operating range / quantity 0.5–50 m; blind zone below 0.5 m Error / dispersion Manufacturer precision 6 cm below 1 m, 4 cm above 1 m; no demonstrated control accuracy Interface UART or CAN, observe ordered variant Specifications, availability & sources Rate / points 100 Hz Weight / housing 110 g · IP67 with applied sealant according to manufacturer; independent IP67 certification not yet completed according to FAQ Supply / power 5–13 V DC; 2 W at 5 V; 5.5 V recommended Field of view / coverage 43° × 30° 0.50 m is the minimum range. No evidence of 0.50 m ±0.05 m on wet glass or in closed-loop flight control. Availability when checked Manufacturer shop lists price and cart but also “Coming Soon”; immediate availability unconfirmed. Purchase / region Ships from USA; manufacturer states EAR99 / NLR and ITAR-free. This does not promise delivery to every country; check local 24 GHz use. - [Ainstein US-D1: technical data, export and application guidance ↗](https://ainstein.ai/us-d1-all-weather-radar-altimeter/) - [Ainstein US-D1 manufacturer shop / price ↗](https://sensing.ai/collections/ainstein-radar-products/products/us-d1radar-altimeter) - [ArduPilot US-D1 integration ↗](https://docs.ardupilot.org/copter/docs/common-aerotenna-usd1.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://sensing.ai/collections/ainstein-radar-products/products/us-d1radar-altimeter) Single-point ToF / LiDARCompare: LightWare SF30/D · 200 m LightWare ### SF30/D · 200 m Long-range height and obstacle channel 459 USDManufacturer price explicitly excludes taxes; shipping and import extra. Supplementary height channel or narrow beam for longer distances; oversized for close-range precision. Assessment at 50 cm sensor distance: 50 cm lies within range; ±5 cm at low or ±10 cm at high rate is insufficient as overall verification of 50 cm ±5 cm. Primarily classify as a height/backup channel. Surface distanceHeight above ground #### Strengths - 200 m nominal range at only 30 g without cable. - Very narrow beam and configurable rate; first/last returns available. #### Weaknesses - High measurement rate comes with ±10 cm nominal error and does not help the ±5 cm goal. - Rear only IP40; housing integration with heat dissipation and EMI protection required. IntegrationMedium Check supported UAV-controller mode; slow filters increase delay, and 20 kHz is not a promised control-loop rate. Operating range / quantity 0.20–200 m; 70% albedo in sunlight, 10 km visibility, 0.9 × 0.9 m target. Error / dispersion ±5 cm below 500 readings/s; ±10 cm above 500 readings/s. The boundary at exactly 500/s is unclear in the guide. Interface UART, I²C; USB for configuration Specifications, availability & sources Rate / points 39–20,000 readings/s; select rate and accuracy together Weight / housing 30 g without cable · Complete device IP40; front IP67 Supply / power 4.5–5.5 V; 96 mA typical, <220 mA at startup Field of view / coverage Beam divergence <0.2° No complete map without additional scan motion and precise pose timestamps; measuring water does not establish performance on wet glass. Availability when checked Manufacturer: US$459 excluding taxes; “Available on backorder”. Purchase / region International manufacturer or retailer ordering; check delivery, import and final tax for Germany before ordering. - [LightWare: SF30/D shop and price ↗](https://lightwarelidar.com/shop/sf30-d-200-m/) - [LightWare: product guide 3.3, 16 March 2026 ↗](https://lightwarelidar.com/wp-content/uploads/2026/03/SF30D-Product-Guide-v3.3-1.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://lightwarelidar.com/shop/sf30-d-200-m/) Industrial 2D scannerCompare: Hokuyo UST-20LX Hokuyo ### UST-20LX Fast Ethernet contour scan for local maps and localisation. 2,758.80 EURRobotShop EU, SKU RB-Seu-11, including taxes; shipping extra. Lightweight industrial scanner for robust contour data; direct sunlight and wet optics remain limitations. Mapping pipeline: 270° contour in one plane. SLAM and motion create larger 2D maps; 3D requires measured motion/orientation. Not a complete flight system. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - 40 Hz scanning at low sensor weight - Ethernet and available ROS1/ROS2 integration #### Weaknesses - 90° unscanned sector and no vertical coverage - Manufacturer requires avoiding direct sunlight despite ambient-light specification IntegrationMedium Ethernet computer, 10–30 V supply, mounting/time calibration and external SLAM/pose processing. Operating range / quantity 0.06–20 m on white Kent paper; up to 8 m at 10% reflectivity; 60 m is only the maximum detection distance Error / dispersion ±40 mm; repeatability σ<30 mm under manufacturer test conditions Interface Ethernet 100BASE-TX; synchronisation output Specifications, availability & sources Rate / points 40 scans/s (25 ms); 0.25° angular step Weight / housing 130 g without cable · IP65 Supply / power 10–30 V; ≤150 mA at 24 V (≤3.6 W), startup about 450 mA Field of view / coverage 270° horizontal, one scan plane 20 m is the specified bright-target range; 60 m is not a guaranteed working range. Not a safety laser scanner and no glass guarantee. Availability when checked Currently in Hokuyo range; RobotShop EU offers the exact UST-20LX version on request. Purchase / region Hokuyo international and RobotShop EU; confirm lead time. - [Hokuyo UST-10/20LX manufacturer specifications ↗](https://www.hokuyo-aut.jp/search/single.php?serial=167) - [Hokuyo USA UST-20LX, ROS and applications ↗](https://www.hokuyo-usa.com/products/lidar-sensors/ust-20lx) - [RobotShop EU UST-20LX, price including taxes ↗](https://eu.robotshop.com/de/products/hokuyo-ust-20lx-scan-laser-entfernungsmesser-eu) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://eu.robotshop.com/de/products/hokuyo-ust-20lx-scan-laser-entfernungsmesser-eu) Very light 2D scannerCompare: LDROBOT STL-27L – Waveshare Kit 24782 LDROBOT ### STL-27L – Waveshare Kit 24782 Budget option for lightweight 360° contour scans and mapping prototypes. 167.99 USDWaveshare kit price SKU 24782; check tax/import status and shipping separately. Very light for dry bench/flight prototypes; wet cleaning conditions require a suitable solution. Mapping pipeline: Planar distance scan. Continuous 2D maps require SLAM and motion; 3D additionally requires known scan-plane orientation/motion. Not a complete navigation system. Mapping / SLAMSurface / object recognition #### Strengths - Very low sensor mass and power consumption - 25 m class and 21.6 kHz sampling in an affordable kit #### Weaknesses - No demonstrated water protection - Single plane requires poses and motion to create a map IntegrationMedium Add UART driver, computer/logger, power supply and calibrated time/pose reference; the ESP8266 adapter is not a SLAM computer. Operating range / quantity 0.03–25 m at about 80% reflectivity Error / dispersion Mean error: ±15 mm up to 2 m; ±20 mm at 2–8 m; ±30 mm beyond Interface UART 921,600 bit/s; kit with ESP8266 adapter and USB TTL Specifications, availability & sources Rate / points 21,600 readings/s; 10 Hz default (6–13 Hz according to wiki) Weight / housing 46.1 g sensor according to Waveshare; kit heavier · Manufacturer datasheet states IP5X; no demonstrated water protection Supply / power 5 V, about 290 mA / 1.45 W Field of view / coverage 360° horizontal, one scan plane Advertised glass detection applies only at certain angles; no reliable glass protection. 25 m is a laboratory figure on bright targets. Availability when checked Waveshare STL27L kit listed; check shop lead time. Do not confuse with D500 or the older LD19/D300 kit. Purchase / region LDROBOT international; Waveshare direct shop, import costs depend on destination. - [LDROBOT current STL-27L sensor range ↗](https://www.ldrobot.com/sensors) - [LDROBOT STL-27L manufacturer datasheet, hosted by morpheusTEK ↗](https://morpheustek.com/wp-content/uploads/2023/09/LDROBOT_STL-27L_Datasheet_EN_v0.3.pdf) - [Waveshare STL27L kit SKU 24782, price and data ↗](https://www.waveshare.com/product/dtof-lidar-stl27l.htm) - [Waveshare STL27L wiki ↗](https://www.waveshare.com/wiki/DTOF_LIDAR_STL27L) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.waveshare.com/product/dtof-lidar-stl27l.htm) 2D scanning LiDARCompare: LightWare SF45/B (50 m) LightWare ### SF45/B (50 m) Horizontal surroundings and sector measurements for obstacle distance 519 USDManufacturer shop explicitly excludes taxes; shipping and import duties extra. Lightweight 2D surroundings sensing to supplement local frontal distance. Mapping pipeline: Planar scan: a continuous 2D map requires motion and poses/SLAM; a spatial map additionally requires known scan-plane motion and orientation. Surface / object recognitionMapping / SLAM #### Strengths - Large adjustable sector at low weight. - First and last returns plus ArduPilot/PX4 support are documented. #### Weaknesses - One scan plane provides no 3D coverage; account for movement during scanning. - Specular targets can cause wrong or missing readings; heat dissipation, shielding and protective housing are required. IntegrationMedium Configure scan sector, masks for aircraft components and filter latency; size supply for startup. Operating range / quantity 0.2–50 m; maximum range specified with a 70% reflective target Error / dispersion ±5 cm below 500 readings/s; ±10 cm above 500 readings/s according to product guide Rev. 3.1 Interface UART, I²C; USB for setup/configuration Specifications, availability & sources Rate / points Up to 5,000 distances/s; up to 5 sweeps/s, depending on configuration Weight / housing 59 g without cable according to current guide; shop text additionally gives older 58 g figure · OEM module; full splash-water IP rating unconfirmed in reviewed documents Supply / power 5–5.5 V; 300 mA typical, <450 mA at startup Field of view / coverage Configurable scan plane up to 320°; one oscillating beam Datasheet ±5 cm at low data rate is not demonstrated ±5 cm flight control or independent glass protection. Availability when checked Manufacturer shop: available on backorder, no confirmed delivery date. Purchase / region Manufacturer direct shop verified; destination and export checks required, LightWare terms list regional delivery restrictions. - [LightWare SF45/B manufacturer shop, price and current key data ↗](https://lightwarelidar.com/shop/sf45-b-50-m/) - [LightWare SF45/B product guide Rev. 3.1, 16 March 2026 ↗](https://lightwarelidar.com/wp-content/uploads/2026/03/SF45B-Product-Guide-v3.1.pdf) - [LightWare pricing, delivery and export terms ↗](https://lightwarelidar.com/terms-and-conditions/) - [PX4 LightWare sensor integration ↗](https://docs.px4.io/main/en/sensor/sfxx_lidar) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://lightwarelidar.com/shop/sf45-b-50-m/) IP67 industrial 2D scannerCompare: SICK TiM571-2050101 (1075091) SICK ### TiM571-2050101 (1075091) Weather-protected 270° distance scanner for contour maps and localisation. 2,015.96 EURRS Germany guide price for 1 unit, excluding VAT; €2,398.99 including VAT; shipping separate. Robust industrial candidate for local contours; weigh mass and 15 Hz rate against lighter alternatives. Mapping pipeline: Provides a planar 270° scan. Contour maps result from registration using SLAM/poses. Spatial maps require motion and known scan-plane orientation. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - Manufacturer specifies indoor/outdoor operation and IP67 - Industrial connectors and Ethernet measurements #### Weaknesses - 25 m drops to 8 m for dark 10% targets - Only one plane; 15 Hz scanning limits fast movement IntegrationMedium Add Ethernet computer and processing; USB service port does not replace the measurement data path. SLAM requires poses/IMU. Operating range / quantity 0.05–25 m; 8 m at 10% remission Error / dispersion Systematic ±60 mm; statistical <20 mm typical (90% target) Interface Ethernet TCP/IP; Micro-USB for service/parameter configuration; PNP synchronisation output Specifications, availability & sources Rate / points 15 Hz; 0.33° angular resolution; typically 67 ms per scan Weight / housing 250 g without cable · IP67 with auxiliary-port cover closed Supply / power 9–28 V DC; typically 4 W Field of view / coverage 270° horizontal, one scan plane Measurement scanner without established suitability as a sole safety function; practically validate glass/spray mist and cleaning agents. Availability when checked Current manufacturer datasheet dated 4 August 2026; RS states manufacturer stock with dispatch from 30 November 2026. Purchase / region SICK/RS sales in Germany/EU and internationally. - [SICK TiM571-2050101 manufacturer datasheet ↗](https://www.sick.com/media/pdf/4/44/444/dataSheet_TiM571-2050101_1075091_en.pdf) - [RS Germany TiM571-2050101 / 1075091, unit price ↗](https://de.rs-online.com/web/p/lichterfassung-und-entfernungsmessung/2133822) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://de.rs-online.com/web/p/lichterfassung-und-entfernungsmessung/2133822) 2D triangulation scannerCompare: SLAMTEC RPLIDAR A3 (A3M1) SLAMTEC ### RPLIDAR A3 (A3M1) Established 360° scanner for 2D SLAM and geometric development experiments. 599 USDDFRobot unit price, SKU DFR0583; check tax/import status and shipping separately. Useful for dry, controlled mapping experiments; less suitable for wet facade work. Mapping pipeline: Planar contour map after registering multiple scans with estimated poses. Spatial reconstruction requires controlled scan-plane motion and registration with IMU/poses. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - Open SDK and broad ROS integration - Switchable indoor/outdoor sampling #### Weaknesses - Percentage distance error grows with distance - No demonstrated spray-water protection in reviewed specifications IntegrationMedium Computer, SLAM software, clean time/pose reference and mechanical protection are additionally required. Operating range / quantity From 0.2 m; up to 25 m on white targets in Enhanced mode, 20 m in Outdoor mode Error / dispersion Manufacturer: 1% up to 3 m; 2% at 3–5 m; 2.5% at 5–25 m Interface TTL UART (3.3 V levels), 256,000 bit/s Specifications, availability & sources Rate / points 16,000 readings/s Enhanced, 10,000 Outdoor; typically 10 Hz (5–15 Hz) Weight / housing 190 g · No IP rating stated in reviewed manufacturer specifications Supply / power 5 V; 450–600 mA, 2.25–3 W Field of view / coverage 360° horizontal, one scan plane Observe the single scan plane, 0.2 m near-range limit and 0–40 °C manufacturer range; no glass safety evidence. Availability when checked Current SLAMTEC range; DFRobot listing with public purchase price. Purchase / region International via DFRobot and SLAMTEC retailers. - [SLAMTEC A3 manufacturer specifications ↗](https://www.slamtec.com/en/lidar/a3spec) - [SLAMTEC A3 authorised distribution, mass ↗](https://www.symphotony.com/lidar/slamtec/rplidar-a3/) - [DFRobot A3, SKU DFR0583, public price ↗](https://www.dfrobot.com/product-1772.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.dfrobot.com/product-1772.html) 2D all-round scannerCompare: SLAMTEC RPLIDAR S2 (S2M1-R2) SLAMTEC ### RPLIDAR S2 (S2M1-R2) Affordable 360° scanning for local contour maps and SLAM prototypes. 399 USDSeeed unit price, explicitly excluding VAT; shipping and import extra. Affordable entry into local 2D mapping; incomplete on its own for free-flying 3D navigation. Mapping pipeline: Produces planar distance contours. A continuous map requires motion and estimated poses/SLAM; a 3D map additionally requires motion with a precise orientation reference. No factory-provided autonomous navigation. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - Dense 360° contour at a manageable price - Manufacturer SDK and ROS/ROS2 drivers available #### Weaknesses - No height information outside the scan plane - Dark targets substantially reduce usable range IntegrationMedium Add companion computer, SDK/ROS, timing reference, mounting calibration and SLAM with pose/IMU data. Operating range / quantity 0.05–30 m at 90%; up to 10 m at 10% reflectivity Error / dispersion ±30 mm; range resolution 13 mm Interface TTL UART, 1 Mbit/s; S2E Ethernet is a different variant Specifications, availability & sources Rate / points 32,000 readings/s; 10 Hz typical Weight / housing 185 g according to manufacturer datasheet; some retailers state 190 g · IP65 Supply / power 5 V; manufacturer states >2 W; account for startup current Field of view / coverage 360° horizontal, one scan plane Test glass, mirrors, spray mist and incidence angle in the actual application; IP65 does not establish approval for cleaning use. Availability when checked Current manufacturer range; Seeed SKU 114992738 publicly listed with unit price. Purchase / region International Seeed distribution; EU purchase possible through retailers. - [SLAMTEC S2 specifications ↗](https://www.slamtec.com/en/s2/spec) - [SLAMTEC S2 datasheet, model S2M1-R2 ↗](https://wiki.slamtec.com/download/attachments/83066883/SLAMTEC_rplidar_datasheet_S2_v2.0_en.pdf?api=v2) - [Seeed S2 product sheet with price, SKU 114992738 ↗](https://files.seeedstudio.com/Bazaar/product_pdf/114992738.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://files.seeedstudio.com/Bazaar/product_pdf/114992738.pdf) Compact 2D all-round scannerCompare: SLAMTEC RPLIDAR S3 SLAMTEC ### RPLIDAR S3 Lighter 360° ToF scanner for local mapping and obstacle contours. 549 USDSeeed unit price; VAT status unclear on product page, check shipping/import separately. Interesting for small SLAM prototypes with a tight weight budget; solve three-dimensional coverage separately. Mapping pipeline: Provides planar 360° points. A larger 2D map requires motion and SLAM/poses; 3D additionally needs known motion and orientation. Not complete flight navigation. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - Lower sensor mass than S2 and A3 - Manufacturer states ≥80 klux ambient-light resistance #### Weaknesses - Single scan plane misses obstacles above and below it - 40 m applies to well-reflecting targets; very dark targets only 5 m IntegrationMedium 5 V supply with startup headroom, computer, driver, timing reference and pose/IMU fusion required. Operating range / quantity 0.05–40 m at 70%; 15 m at 10%; 5 m at 2% reflectivity Error / dispersion ±30 mm; 10 mm resolution Interface TTL-UART, 1 Mbit/s Specifications, availability & sources Rate / points 32,000 readings/s; 10–20 Hz, typically 10 Hz Weight / housing 115 g · IP65 Supply / power 5 V, typically 450 mA (about 2.25 W); startup typically 1.2 A Field of view / coverage 360° horizontal, one scan plane Do not assume reliable glass or facade-distance protection; separately test water film and spray mist. Availability when checked Currently listed by SLAMTEC; Seeed SKU 101090041 listed in stock when checked. Purchase / region International Seeed and SLAMTEC retailer distribution. - [SLAMTEC S3 specifications ↗](https://www.slamtec.com/en/s3/spec) - [SLAMTEC S3 manufacturer datasheet, hosted by Seeed ↗](https://files.seeedstudio.com/wiki/SLAMRadar/S3/datasheet_SLAMTEC_rplidar_datasheet_S3_v1.0_en.pdf) - [Seeed S3, SKU 101090041, public unit price ↗](https://www.seeedstudio.com/RPLiDAR-S3M1-p-5753.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.seeedstudio.com/RPLiDAR-S3M1-p-5753.html) IP65 2D all-round scannerCompare: YDLIDAR / EAI TG30 YDLIDAR / EAI ### TG30 More robust 360° ToF scanning for 2D contours and SLAM. 359 EURExp-Tech SKU 750-TG30, explicitly excluding VAT and plus shipping; B2B shop. 2D SLAM and contour experiments with an enclosed scanner; check version/weight before procurement. Mapping pipeline: Individual planar scans form a registered map only with motion and poses. 3D reconstruction requires additional known scan-plane motion and orientation. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - IP65 and enclosed design - Public EU price including USB adapter #### Weaknesses - Only one plane and varying distance error across range - Older product revisions have different weight/current specifications IntegrationMedium Add YDLIDAR SDK/ROS, companion computer and SLAM/IMU/pose fusion; confirm exact supplied revision. Operating range / quantity 0.05–30 m; account for target reflectivity and lighting Error / dispersion ±60 mm up to 5 m; ±40 mm at 5–20 m; ±100 mm at 20–30 m (80% test target) Interface UART 512,000 bit/s; Exp-Tech package includes USB adapter Specifications, availability & sources Rate / points 20,000 readings/s; 5–12 Hz Weight / housing 140 g net according to manufacturer datasheet V1.4; older retailer documents state 214 g · IP65 Supply / power 5 V (4.8–5.2 V); typically 340 mA, about 1.7 W; startup up to 1 A Field of view / coverage 360° horizontal, one scan plane Avoid direct sunlight into receiver; do not assume reliable detection of glass or spray mist. Availability when checked Current manufacturer page; Exp-Tech listed five units in stock when checked. Purchase / region Ships from Germany via Exp-Tech; other international distribution. - [YDLIDAR TG30 manufacturer page ↗](https://www.ydlidar.com/product/ydlidar-tg30) - [YDLIDAR TG30 manufacturer datasheet V1.4 ↗](https://www.ydlidar.com/Public/upload/files/2022-06-21/YDLIDAR%20TG30%20Data%20Sheet%20V1.4%28211230%29.pdf) - [Exp-Tech TG30, public EU price and package contents ↗](https://exp-tech.de/en/products/tg30) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://exp-tech.de/en/products/tg30) Compact rotating 3D LiDARCompare: Hesai JT128 Hesai ### JT128 Lightweight, dense near-surroundings sensing with a large solid angle. 1,200 EURNet, excluding 19% VAT and shipping according to offer. Interesting lightweight sensor for all-round vision and facade sensing; measure actual coverage on the aircraft frame. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Many channels at low sensor weight support detailed close-surroundings sensing. - Dual return and time synchronisation are documented for a moving platform. #### Weaknesses - Manufacturer marketing and manual use different vertical-field-of-view figures. - Uneven firing order requires correct timestamps and motion compensation. IntegrationMedium Always use the supplied unit’s firetime and angle corrections. Manual requires channel-specific motion compensation in mobile and precision-critical applications. Operating range / quantity 0–60 m instrumented range; 40 m at 10% reflectance across all channels. Error / dispersion Typically ±3 cm accuracy; 2 cm precision (standard deviation) in current J01 manual. Do not equate with complete flight-control error. Interface 100BASE-TX Ethernet; GNSS/PTP, IMU; official Hesai SDK and ROS/ROS2 driver support JT128. Specifications, availability & sources Rate / points 10/20 Hz; 1.152 million points/s single, 2.304 million dual. Weight / housing Under 265 g according to current manual; product page rounds to 265 g. · IP6K7 in current J01 manual. Supply / power 8.4 W typical at 25 °C and 10 Hz, without accessories, according to directly reviewed manufacturer manual J01-en-261010 (manual pp. 17–18, PDF pp. 21–22); 9–32 V DC. Allow at least 30 W available supply. Field of view / coverage 360° horizontal. Current manual: 95° vertical channel angle (−4.4° to +90.5°); product page advertises 189° hyperhemispherical total coverage. Do not equate these without explaining geometry. A 0 m measurement range does not establish contact or glass detection. Resolve the geometry of the 95° channel angle versus 189° advertised coverage during installation planning. Glass, wet surfaces and spray mist are unproven in the cleaning setup; no guaranteed 50 cm ±5 cm nozzle standoff. Power specification is revision-specific: 8.4 W was confirmed in the directly downloaded J01-en-261010; do not transfer differing values from older manuals to this revision. Availability when checked QUADRUPED.DE lists EUR 1,200 net, immediately available and 2–6 working days domestically; international delivery differs. Purchase / region German retailer offer; Hesai sales in Europe, USA and Asia. Check delivery area and returns for the specific offer. - [Hesai: JT128 manual J01-en-261010, directly reviewed; specifications pp. 16–18 (PDF 20–22) ↗](https://www.hesaitech.com/wp-content/uploads/2026/10/JT128_User_Manual_J01-en-261010.pdf) - [Hesai: JT128 product page with advertised 189° figure ↗](https://www.hesaitech.com/product/jt128/) - [QUADRUPED.DE: JT128, net price and stock ↗](https://www.quadruped.de/Hesai-JT128) - [Hesai: official ROS/ROS2 driver with model list ↗](https://github.com/HesaiTechnology/HesaiLidar_ROS_2.0/blob/master/README.md) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.quadruped.de/Hesai-JT128) Compact rotating 3D LiDARCompare: Hesai JT16 Hesai ### JT16 Lightweight all-round sensing as an affordable test sensor. 479 EURIncludes statutory German VAT according to retailer tax section; product page simultaneously shows a conflicting additional tax note. Confirm at checkout. Weight and power consumption are attractive for compact drones. Mounting angle must account for the upward-facing vertical fan. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Very small weight and power budgets simplify an experimental setup. - Public documentation and drivers simplify data acquisition. #### Weaknesses - Only 16 channels and 2.67° vertical sampling provide less wall detail. - 3 cm standard deviation leaves little allowance for motion, calibration errors and control within a ±5 cm total goal. IntegrationMedium Observe serial interface, firmware baud rate and angle-correction file; do not assume connection box and cables are included in the base price. Operating range / quantity 0.15–100 m instrumented measurement range; 30 m at 10% reflectance across all channels, 50% detection probability and 0–100 klx. Error / dispersion Typically ±3 cm accuracy; 3 cm precision (1σ). These separate specifications are not overall SLAM or nozzle-control accuracy. Interface Native RS485 at 3,000,000 baud; certain firmware 3,125,000 baud. Ethernet via suitable connection box; official Hesai SDK and ROS driver list JT16. Specifications, availability & sources Rate / points 5/10 Hz; 48,000 valid points/s according to J03 manual. Weight / housing 199.7 g. · IP6K6; not specified as submersible. Supply / power 4.3 W typical at 14 V / 5 Hz; 12–16 V DC; at least 16 W supply headroom according to manual. Field of view / coverage 360° horizontal × 40° vertical (0° to +40°); average vertical resolution 2.67°. Do not present 100 m as guaranteed target range. Public 96,000 points/s figures differ from the current J03 manual’s 48,000 valid points/s. No demonstrated suitability for glass, water or spray mist, and no guarantee of 50 cm ±5 cm. Availability when checked OpenELAB displays sensor at EUR 479 with units in stock; stock and promotional price can change at short notice. Purchase / region Price offer for Germany; OpenELAB directs international customers to its global shop. Hesai operates European and US sales. - [Hesai: JT16 manual J03 dated 20 January 2026 ↗](https://www.hesaitech.com/wp-content/uploads/2026/01/JT16_User_Manual_J03-en-260120.pdf) - [Hesai: current JT32/JT16 product family ↗](https://www.hesaitech.com/product/jt16/) - [OpenELAB Germany: JT16 price, stock and tax notes ↗](https://openelab.de/products/hesai-jt16-mini-3d-lidar) - [Hesai: official ROS/ROS2 driver with model list ↗](https://github.com/HesaiTechnology/HesaiLidar_ROS_2.0/blob/master/README.md) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://openelab.de/products/hesai-jt16-mini-3d-lidar) 3D LiDARCompare: Hesai XT16 Hesai ### XT16 Industrial comparison Price on requestRequest a current quotation 16-channel sensor with 360° × 30°; for mapping and geometry. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Industrial 360° sensor with documented manufacturer specifications. - Suitable as a comparison for point-cloud quality and geometric measurement. #### Weaknesses - 800 g weight and only 30° vertical field of view burden UAV design. - No confirmed current list price or independent evidence on wet glass. IntegrationHigh Ethernet/SDK/ROS integration on companion computer, calibration and a custom odometry/mapping pipeline required. Operating range / quantity 0.05–120 m Error / dispersion Typically ±1 cm accuracy; 0.5 cm spread (1σ), observe test conditions. Interface Ethernet / manufacturer SDK / ROS Specifications, availability & sources Rate / points 320,000 points/s, single return Weight / housing 800 g · IP6K7 Supply / power No reliable figure in the reviewed excerpt; check manufacturer datasheet before design Field of view / coverage 360° × 30°; 2° vertical angular resolution Allow for 800 g and narrow vertical field of view. No demonstrated maximum error on wet glass at 50 cm. Earlier search prices are not currently confirmed. Availability when checked Request quotation and specific lead time from manufacturer or distributor Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Hesai official XT32M/32/16 specification table ↗](https://www.hesaitech.com/product/xt16-32-32m/) - [Quadruped retailer status ↗](https://www.quadruped.de/Hesai-XT16) - [Hesai XT16 manufacturer manual ↗](https://www.hesaitech.com/wp-content/uploads/2023/03/XT16_User_Manual_X02-en-220410.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.quadruped.de/Hesai-XT16) 3D LiDARCompare: Hesai XT32 Hesai ### XT32 Denser 3D geometry Price on requestNo reliable public new-unit price verified 32 channels, 360° × 31°; finer vertical sampling than XT16. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Denser vertical sampling than XT16. - 360° geometry for industrial mapping and navigation tasks. #### Weaknesses - At 800 g substantially heavier than compact UAV/robotics sensors. - Do not confuse with XT32M or M2X; measurement performance depends on revision and test conditions. IntegrationHigh As for XT16, with greater point-data volume; size processing, bandwidth, synchronisation and calibration. Operating range / quantity 0.05–120 m Error / dispersion Typically ±1 cm accuracy; 0.5 cm spread (1σ), observe test conditions. Interface Ethernet / manufacturer SDK / ROS Specifications, availability & sources Rate / points 640,000 points/s, single return Weight / housing 800 g · IP6K7 Supply / power No reliable figure in the reviewed excerpt; check manufacturer datasheet before design Field of view / coverage 360° × 31°; 1° vertical angular resolution Do not confuse with XT32M. Check weight and operating conditions. Datasheet does not establish reliability on glass or in spray mist. Availability when checked Request quotation and specific lead time from manufacturer or distributor Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Hesai official XT32M/32/16 specification table ↗](https://www.hesaitech.com/product/xt16-32-32m/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.hesaitech.com/product/xt16-32-32m/) Rotating 3D LiDARCompare: Hesai XT32M2X (XT32M family) Hesai ### XT32M2X (XT32M family) Precise 3D geometry, SLAM and surveying. 4,636.40 EURNet, excluding VAT; shipping and project-specific accessories extra. Worth testing for geometric measurements on drones with sufficient payload capacity. A 50 cm sensor-to-surface distance is at the minimum limit; a nozzle offset changes the required sensor distance. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Defined accuracy and precision figures enable a clear technical comparison. - Multiple returns can reveal different surface components. #### Weaknesses - At 50 cm sensor distance, the 0.5 m minimum leaves no downward allowance; account separately for sensor-to-nozzle offset. - 1.3° vertical sampling and mounting position limit surface coverage. IntegrationMedium Implement time synchronisation, motion compensation, sensor pose and separate close-range protection. Do not apply general XT zero-blind-zone advertising to this specific model. Operating range / quantity 0.5–300 m measurement range; 80 m at 10% reflectance across all channels (100 klx, >90% detection probability). Error / dispersion ±1 cm accuracy; 0.5 cm typical precision (1σ), standard value 1 cm. Typical test: 0.5–50 m, 50% reflectance, 30 °C; do not transfer to glass/spray mist. Interface 100BASE-TX Ethernet; GNSS/PTP; official Hesai SDK and ROS/ROS2 driver support XT32M2X. Specifications, availability & sources Rate / points 5/10/20 Hz; 640,000 points/s single, 1.28 million dual, 1.92 million triple. Weight / housing 490 g without additional integration accessories. · IP6K7; Class 1 laser. Supply / power 10 W typical; 9–36 V DC; manufacturer requires at least 30 W available supply power. Field of view / coverage 360° horizontal × 40.3° vertical (−20.8° to +19.5°). 0.5 m is the documented lower measurement-range limit, not an accuracy test distance establishing zero blind zone. 1σ precision is not a guaranteed absolute error; glass/water/spray mist unproven. No proof of 50 cm ±5 cm nozzle standoff. Availability when checked Sensor Partners lists SKU 230-0050 as “In stock” at EUR 4,636.40 net. Confirm dispatch, revision and accessories before ordering. Purchase / region European specialist distribution with international Hesai sales network. - [Hesai: XT32M2X manual X03 dated 10 July 2026 ↗](https://www.hesaitech.com/wp-content/uploads/2026/07/XT32M2X_User_Manual_X03-en-260710.pdf) - [Sensor Partners: XT32M2X, price and lead time ↗](https://sensorpartners.com/en/product/xt-32m/) - [Hesai: official ROS/ROS2 driver with model list ↗](https://github.com/HesaiTechnology/HesaiLidar_ROS_2.0/blob/master/README.md) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://sensorpartners.com/en/product/xt-32m/) 3D LiDAR with internal IMUCompare: Livox Avia Livox ### Avia Directional 3D point clouds for SLAM and custom aerial mapping integration. 1,420 EURGebrüder Kraus: net; €1,689.80 including 19% VAT, shipping from €25. Light enough for custom 3D SLAM prototypes; directional field of view requires suitable mounting and additional coverage. Mapping pipeline: Produces actual spatial points. Continuous maps require LiDAR/IMU SLAM and poses; georeferenced survey results additionally need GNSS/INS and calibrated processing. Not a complete payload. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - 3D point cloud and integrated IMU - Multiple echoes and high point rate for mapping #### Weaknesses - No all-round field of view - Internal IMU alone does not produce a georeferenced survey map IntegrationHigh Add computer, time synchronisation, LiDAR/IMU calibration and SLAM; for surveying, add GNSS/INS, trajectory and processing software. Operating range / quantity At 100 klux: 190 m at 10%, 320 m at 80%; 450 m at 80% only at 0 klux Error / dispersion Distance precision 2 cm (1σ at 20 m on an 80% target), not global map accuracy Interface 100 Mbit Ethernet; PTPv2/PPS/GPS time; internal BMI088 IMU Specifications, availability & sources Rate / points 240,000 points/s single; 480,000 dual; 720,000 triple return Weight / housing 498 g without cable · IP67 sensor; converter and cables excluded Supply / power 8 W non-repetitive / 9 W repetitive, startup 16 W; 10–15 V (Converter 2.0: 9–30 V) Field of view / coverage Non-repetitive 70.4° × 77.2°; repetitive 70.4° × 4.5° 450 m is a dark-condition/bright-target figure. Do not equate point/distance precision with map or facade-distance accuracy; no glass evidence. Availability when checked Still listed on official Livox site; German retailer states in-stock goods and 48-hour dispatch. Purchase / region German retailer with EU shipping; international Livox distribution. - [Livox Avia official specifications ↗](https://www.livoxtech.com/avia/specs) - [Livox Avia manufacturer page ↗](https://www.livoxtech.com/avia) - [Gebrüder Kraus Livox Avia, public net price ↗](https://www.livoxtech.de/products/avia.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.livoxtech.de/products/avia.html) 3D LiDARCompare: Livox MID-360 Livox ### MID-360 Existing systems / mapping Not orderable from TPINo current purchase price from the reviewed retailer Reference for existing systems; predecessor to MID-360S. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Lightweight 360° sensor with an established SDK/ROS data path. - Documented as a reference and for existing mapping systems. #### Weaknesses - Previous generation; no longer orderable from reviewed retailer. - Close range and certain reflective or water-like surfaces have explicit limitations. IntegrationHigh Companion computer, SDK/driver, time synchronisation, IMU/sensor extrinsics and a custom LIO/SLAM pipeline required. Operating range / quantity 0.10–70 m, target dependent Error / dispersion ≤3 cm spread (1σ) at 0.2 m; ≤2 cm at 10 m. Interface Ethernet / Livox SDK2 / ROS Specifications, availability & sources Rate / points 10 Hz typical; 200,000 points/s Weight / housing 265 g · IP67 Supply / power 6.5 W average; 9–27 V Field of view / coverage 360° horizontal; vertical −7° to 52° No longer orderable from reviewed retailer. Near range 0.1–1 m has surface caveats; maximum error cannot be derived from 1σ. Availability when checked No longer orderable from reviewed retailer TPI Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Livox official technical data ↗](https://www.livoxtech.com/mid-360/specs) - [TPI retailer status for previous generation ↗](https://tpi-livox.eu/products/mid360.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://tpi-livox.eu/products/mid360.html) 3D LiDARCompare: Livox MID-360S Livox ### MID-360S Surroundings / SLAM 599 EURConfirmed net price; taxes and shipping are additional. Affordable candidate for 3D mapping and local motion estimation. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Price-accessible 360° 3D sensor at 265 g. - Ethernet and SDK2/ROS2 as a basis for mapping and odometry. #### Weaknesses - ≤4 cm at 0.2 m is 1σ spread, not a guaranteed maximum error. - Observe surface caveats at 0.1–1 m and limited vertical coverage. IntegrationHigh Integrate a current model-specific driver, companion computer, synchronisation and LIO/SLAM processing; flight controller receives derived measurements/pose. Operating range / quantity 0.10–100 m; 40 m at 10% reflectance Error / dispersion ≤4 cm spread (1σ) at 0.2 m; ≤2 cm at 10 m. Interface Ethernet / SDK2 / ROS 2 Specifications, availability & sources Rate / points 10 Hz typical; 200,000 points/s Weight / housing 265 g · IP67 Supply / power 6.5 W; 9–27 V DC Field of view / coverage 360° horizontal; vertical −7° to +52° Manufacturer lists restrictions at 0.1–1 m for certain surfaces. 1σ is not a guaranteed ±4 cm bound. Provide independent close-range sensing. Availability when checked Authorized EU dealer offers ordering; confirm lead time. Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Livox official technical data ↗](https://www.livoxtech.com/mid-360s/specs) - [TPI retailer net price ↗](https://tpi-livox.eu/products/mid360s.html) - [Gebrüder Kraus German retailer price ↗](https://www.livoxtech.de/products/mid-360.html) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://tpi-livox.eu/products/mid360s.html) 3D LiDARCompare: Ouster OS0-128 Rev8 Ouster ### OS0-128 Rev8 Premium / industrial comparison 14,923 EURIncludes 19% VAT; shipping extra; special order 128 channels, 360° × 90°, broad 3D field of view and industrial housing. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Broad 360° × 90° coverage and dense vertical sampling. - Robust housing and documented SDK/ROS ecosystem. #### Weaknesses - High purchase price, weight and power demand compared with compact sensors. - Quoted accuracy applies only beyond 1 m; no evidence for 50 cm nozzle standoff. IntegrationHigh Gigabit Ethernet, adequate power supply and companion processing; check revision, scan mode, synchronisation and calibration. Operating range / quantity 35 m at 10% and 75 m at 80% reflectance (100 klx, >90% detection probability); 500 m is the encodable range. Minimum distance 0.5 m default, optional 0.3/0.0 m. Error / dispersion Typically ±1.25 cm on diffuse targets and ±2.5 cm on retroreflectors; precision 0.25–1.5 cm (1σ). These figures apply beyond 1 m at 1024 × 10 Hz. 1 mm distance resolution is not an accuracy guarantee. Interface UDP over Gigabit Ethernet; HTTPS API, synchronous IMU, PTP/gPTP/PPS; manufacturer SDK plus ROS and ROS2. Specifications, availability & sources Rate / points 5–40 Hz depending on mode; 4096 horizontal samples at 5/10 Hz; up to 10.49 million returns/s. Weight / housing 500 g without cap/baseplate; 580 g with cap. · IP68 / IP69K with I/O cable connected and correctly locked. Supply / power 15 W nominal, 10–20 W at 1024 × 10 Hz; supply at least 30 W. Field of view / coverage 360° horizontal × 90° vertical (−45° to +45°). Accuracy specifications apply beyond 1 m under stated test conditions. Adjustable minimum distance does not establish accuracy at 50 cm. Glass, wetting and spray mist require testing in the actual setup. Availability when checked Retailer special order; confirm lead time Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked supplier - [Ouster official OS0 product ↗](https://ouster.com/products/hardware/os0-lidar-sensor) - [Ouster OS0 Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os0.pdf) - [Reichelt exact OS0-128 Rev8 retailer price ↗](https://www.reichelt.com/de/en/shop/product/ouster_multi-layer_3d_lidar_os0_rev8_128_layers_90_0_-_250_m-441657) - [Ouster: current Rev8 product family ↗](https://ouster.com/rev8) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.reichelt.com/de/en/shop/product/ouster_multi-layer_3d_lidar_os0_rev8_128_layers_90_0_-_250_m-441657) Rotating 3D LiDARCompare: Ouster OS1 Max Rev8 (64 / 128 / 256 channels) Ouster ### OS1 Max Rev8 (64 / 128 / 256 channels) Long-range 3D sensing and large-scale facade geometry. Price on requestPrice on request; taxes and shipping depend on quotation. More suited to larger payload platforms and surveying before cleaning. Long range alone does not improve 50 cm nozzle standoff. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - High point density and range support modelling larger buildings. - Native colour and integrated IMU simplify integration into a perception pipeline. #### Weaknesses - Higher weight and energy demand than OS1 and compact robotics LiDAR. - Extra range provides no demonstrated benefit for pure close-range control. IntegrationMedium Allow for companion computer, thermally suitable mount, time synchronisation and calibrated sensor-to-nozzle transformation. Integrate point cloud into flight and distance estimation. Operating range / quantity 200 m at 10% and 350 m at 80% reflectance (100 klx, >90% detection probability); 500 m maximum encodable range. Minimum range 0.5 m default; optional 0.3/0.0 m. Error / dispersion Typically ±1.25 cm on diffuse targets and ±2.5 cm on retroreflectors; precision 0.25–1.5 cm (1σ). These figures apply beyond 1 m at 1024 × 10 Hz. 1 mm distance resolution is not an accuracy guarantee. Interface UDP over Gigabit Ethernet; HTTPS API, synchronous IMU, PTP/gPTP/PPS; manufacturer SDK plus ROS and ROS2. Specifications, availability & sources Rate / points 5–40 Hz depending on configuration; Max has 4096 horizontal samples at 5 Hz. Up to 10.49 million returns/s. Weight / housing 670 g without cap/baseplate; 720 g with cap. · IP68 / IP69K with I/O cable connected and correctly locked. Supply / power 19 W nominal, 15–25 W at 1024 × 10 Hz; allow at least 30 W supply power. Field of view / coverage 360° horizontal × 43.9° vertical. The current Rev8 long-range model is called OS1 Max. OS2 Rev8 is not established on the current manufacturer page. No evidence for reliable glass or spray-mist measurements, and no guarantee of 50 cm ±5 cm nozzle standoff. Availability when checked Currently offered as Rev8 and available for enquiry from manufacturer. Launch announcement states order availability and planned shipping; confirm exact variant and lead time in quotation. Purchase / region International manufacturer sales; regional distributors. Obtain firm availability confirmation for the selected variant in Germany. - [Ouster OS1 Max Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os1-max.pdf) - [Ouster: current OS1 Max product ↗](https://ouster.com/products/hardware/os1-max) - [Ouster: current Rev8 product family ↗](https://ouster.com/rev8) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://ouster.com/rev8) Rotating 3D LiDARCompare: Ouster OS1 Rev8 (32 / 64 / 128 channels) Ouster ### OS1 Rev8 (32 / 64 / 128 channels) Medium range for facade models and obstacle detection. Price on requestPrice on request; taxes and shipping depend on quotation. Plausible as a 3D sensor on a drone with sufficient payload capacity. The test specification beyond 1 m does not establish ±5 cm nozzle control at 50 cm. Mapping pipeline: Spatial point clouds become a map through registration, time-aligned IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Native colour points can support identification of facade sections. - 360° point cloud provides environmental context beyond a single distance value. #### Weaknesses - Power demand, mounting and processing load matter on small drones. - 44° vertical coverage requires mounting suited to the facade. IntegrationMedium Allow for companion computer, thermally suitable mount, time synchronisation and calibrated sensor-to-nozzle transformation. Integrate point cloud into flight and distance estimation. Operating range / quantity 90 m at 10% and 170 m at 80% reflectance (100 klx, >90% detection probability); 500 m maximum encodable range. Minimum range 0.5 m default; optional 0.3 or 0.0 m. Error / dispersion Typically ±1.25 cm on diffuse targets and ±2.5 cm on retroreflectors; precision 0.25–1.5 cm (1σ). These figures apply beyond 1 m at 1024 × 10 Hz. 1 mm distance resolution is not an accuracy guarantee. Interface UDP over Gigabit Ethernet; HTTPS API, synchronous IMU, PTP/gPTP/PPS; manufacturer SDK plus ROS and ROS2. Specifications, availability & sources Rate / points 5–40 Hz depending on mode; 4096 horizontal samples at 5/10 Hz. Up to 10.49 million returns/s, not in every combination. Weight / housing 500 g without cap/baseplate; 580 g with cap. · IP68 / IP69K with I/O cable connected and correctly locked. Supply / power 15 W nominal, 10–20 W at 1024 × 10 Hz; size supply for at least 30 W. Field of view / coverage 360° horizontal × 44° vertical (−22° to +22°). 500 m means representable distance, not assured detection. Glass, wet surfaces and spray mist are unproven for this cleaning setup; test close range separately. Designed for functional safety does not mean all certifications are complete or that the drone is certified. Availability when checked Currently offered as Rev8 and available for enquiry from manufacturer. Launch announcement states order availability and planned shipping; confirm exact variant and lead time in quotation. Purchase / region International manufacturer sales; regional distributors. Obtain firm availability confirmation for the selected variant in Germany. - [Ouster OS1 Rev8: manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-os1.pdf) - [Ouster: current Rev8 product family ↗](https://ouster.com/rev8) - [Ouster: Rev8 availability and context ↗](https://ouster.com/insights/blog/introducing-rev8-os-family) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://ouster.com/rev8) 3D spinning LiDARCompare: Ouster OSDome Rev8 (32 / 64 / 128 channels) Ouster ### OSDome Rev8 (32 / 64 / 128 channels) Hemispherical near-field sensing and obstacle context. Price on requestPrice on request; taxes and shipping depend on the quotation. Interesting for broad near-field sensing if propellers, frame and hose do not obstruct the field of view. Mapping pipeline: Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Hemispherical sensing can cover several obstacle directions. - Native colour and shared Ouster tools standardize data processing. #### Weaknesses - A broad field of view generates many points that are irrelevant to nozzle control. - Precision varies with target and distance; the best value is not a general error bound. IntegrationMedium Allow for a companion computer, thermally suitable mount, time synchronization and calibrated sensor/nozzle transformation. Integrate the point cloud into flight and distance estimation. Operating range / quantity 20 m at 10% and 45 m at 80% reflectivity (100 klx, >90% detection probability). Datasheet: 500 m encodable; minimum range 0.5 m standard, optionally 0.3/0.0 m. Error / dispersion Typically ±1.25 cm on diffuse targets / ±2.5 cm on retroreflectors; precision 0.5–5 cm (1σ), beyond 1 m at 1024 × 10 Hz. 1 mm resolution is not accuracy. Interface UDP over Gigabit Ethernet; HTTPS API, synchronous IMU, PTP/gPTP/PPS; manufacturer SDK plus ROS and ROS 2. Specifications, availability & sources Rate / points 5–40 Hz depending on mode; 4096 samples at 5/10 Hz; up to 10.49 million returns/s. Weight / housing 540 g without baseplate. · IP68 / IP69K with the I/O cable connected and correctly locked. Supply / power 15 W nominal; 10–20 W, plus allowance for accessories and startup. Field of view / coverage 360° horizontal × 180° vertical; hemispherical coverage according to the manufacturer. The manufacturer product page and datasheet differ in their maximum numeric range; the datasheet figure is described here as encodable range. 180° coverage and IP protection do not demonstrate measurement stability through spray. Validate 50 cm nozzle control separately. Availability when checked Currently offered as Rev8 and available for quotation from the manufacturer. The launch announcement gives ordering availability and planned shipping; confirm the specific variant and lead time in the quote. Purchase / region International manufacturer sales and regional distribution partners. Obtain firm confirmation of availability of the chosen variant for Germany. - [Ouster OSDome Rev8 — manufacturer datasheet v4.0 ↗](https://data.ouster.io/downloads/datasheets/datasheet-rev8-v4p0-osdome.pdf) - [Ouster — current Rev8 product family ↗](https://ouster.com/rev8) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://ouster.com/rev8) 3D LiDARCompare: RoboSense Airy · 96 Beam RoboSense ### Airy · 96 Beam Broad environmental perception 999 USDDisplayed shop price; tax status unconfirmed. Check shipping and import costs separately. Lightweight 3D comparison candidate with a 360° × 90° field of view. Mapping pipeline: Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Broad 360° × 90° field of view at low sensor weight. - Industrial enclosure and documented Ethernet/ROS connection. #### Weaknesses - Check field-of-view orientation and the mode-specific scan gap. - 1σ precision is not maximum flight deviation; confirm the German checkout price and delivery date. IntegrationHigh ROS driver, scan mode, synchronization, calibration and companion processing must match the flight and map coordinate frames. Operating range / quantity 0.10–60 m; 30 m at 10% reflectivity Error / dispersion Typically 1.5 cm dispersion (1σ), using a specified diffuse reflector. Interface Ethernet / UDP / ROS Specifications, availability & sources Rate / points 10 Hz; 856,320 points/s Weight / housing 230 ±20 g · IP67 / IP6K9K Supply / power 8 W typical; 9–32 V DC Field of view / coverage 360° horizontal; 0° to 90° vertical 1σ is not a maximum error bound. Check orientation of the 0–90° field and the documented scan gap in 96-beam mode; confirm availability. Availability when checked Manufacturer shows availability notification; lead time unconfirmed. Purchase / region International manufacturer/distributor catalogue; confirm delivery to Germany with the linked vendor - [RoboSense official shop: 96-beam price ↗](https://store.robosense.ai/products/airy?variant=44401434165301) - [RoboSense official manual/wiki ↗](https://robosense-robotics.github.io/product-manual/en/Airy/) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.robosense.ai/products/airy?variant=44401434165301) Compact digital 3D spinning LiDARCompare: RoboSense Fairy 48-Beam – Device Only RoboSense ### Fairy 48-Beam – Device Only Lightweight sensor for dense 3D façade points and UAV mapping. 1,699 USDManufacturer shop mentions DDP and duties/import taxes calculated at checkout. Check the destination-specific final amount; accessory kit is separate. Low weight and a structured point cloud are promising for UAV trials. Resolve missing near-range and variant data before selection. Mapping pipeline: Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - High point count below 350 g is attractive for a drone. - The manufacturer explicitly lists drones as an application. #### Weaknesses - Do not mix 48/96-channel specifications or accessory variants. - Advertised 1σ precision does not establish absolute accuracy on glass or at 50 cm. IntegrationMedium Select the correct TX/T1 connection, firmware, IMU calibration and driver version. First establish the offered unit’s supply, protection rating and minimum distance. Operating range / quantity Up to 150 m manufacturer maximum; the reviewed product page does not separately establish minimum detection distance. 0.5 m is the lower limit of a published precision test, not necessarily the blind zone. Error / dispersion Manufacturer states 0.5 cm (1σ): classify as precision, not guaranteed ±0.5 cm absolute error. Family page: 0.5–70 m, 50% reflectivity, 30 °C, channels 25–72; confirm applicability to the supplied 48-channel variant. Interface RoboSense SDK with ROS/ROS 2 support lists RS-LiDAR-FAIRY; TX/T1 accessory variants in the shop, so specify the connection standard. Specifications, availability & sources Rate / points 48-channel shop version: approximately 690,000 points/s single and 1.38 million dual; horizontal/vertical sampling 0.25° × 0.67°. Weight / housing Under 350 g according to manufacturer; φ75 × 70 mm. · Not fully specified for the exact shop variant on the reviewed manufacturer pages; request a binding variant datasheet. Supply / power No verified power figure for the exact 48-channel variant on the reviewed manufacturer pages; request a binding datasheet. Field of view / coverage 360° horizontal × 32° vertical. Do not derive near-range suitability from the 0.5 m test value. 1σ is neither maximum deviation nor a 50 cm ±5 cm system guarantee. Glass, wet façades and cleaning spray remain unproven. Availability when checked Listed with a price in the manufacturer shop, currently with an availability notification rather than confirmed immediate stock. European retailer ROS Components quotes ordering with 4–6 weeks lead time. Purchase / region International manufacturer shop and European specialist retailers; confirm destination delivery and stock. - [RoboSense — Fairy family and test conditions ↗](https://www.robosense.ai/en/rslidar/Fairy) - [RoboSense — public price for the exact 48-channel version ↗](https://store.robosense.ai/products/fairy-48-beam-3d-lidar) - [ROS Components — Fairy European backorder listing ↗](https://www.roscomponents.com/es/producto/fairy/) - [RoboSense — official SDK with ROS/ROS 2 support ↗](https://github.com/RoboSense-LiDAR/rslidar_sdk) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.robosense.ai/products/fairy-48-beam-3d-lidar) 32-channel 3D surround LiDARCompare: RoboSense Helios 32 (H32F70) RoboSense ### Helios 32 (H32F70) Three-dimensional surround view with strong downward coverage for mapping/SLAM. Price on requestExact H32F70 unit price is unclear in the multi-variant shop; request a Device Only quotation. Strong 3D surround mapping on a sufficiently capable platform; about 1 kg of sensor plus wiring is substantial for a small drone. Mapping pipeline: Produces spatial surround point clouds. Mapping needs a trajectory/poses, IMU, synchronization and a SLAM computer; georeferencing also needs GNSS/INS. Not a complete flight-navigation package. Mapping / SLAMSurface / object recognitionPosition / navigation #### Strengths - True spatial 360° data rather than a single scan plane - 70° vertical field focused below the sensor #### Weaknesses - High mass compared with compact solid-state sensors - 100BASE-T1 may require a suitable interface box/media converter IntegrationHigh Add RoboSense drivers, a computer, Ethernet connection, synchronized IMU/poses and LiDAR SLAM; specify the connection kit. Operating range / quantity Up to 150 m; 110 m at 10% NIST reflectivity; blind zone ≤0.2 m Error / dispersion Typical range precision 1 cm on a 50% NIST target, dependent on channel/environment Interface 100BASE-T1; UDP; PPS/$GPRMC, PTP/gPTP Specifications, availability & sources Rate / points 5/10/20 Hz; 576,000 points/s single, 1,152,000 dual return Weight / housing 1,000 ±50 g · IP67 and IP6K9K according to current shop Supply / power 9–32 V; 12 W at 10 Hz Field of view / coverage 360° horizontal × 70° vertical (−55° to +15°) Do not mix current H32F70 data with older RS-Helios-16P data. IP6K9K is not blanket approval for cleaning chemicals or glass navigation. Availability when checked H32F70 currently listed in the official RoboSense shop; order through customer service and confirm availability. Purchase / region Official international RoboSense shop; check destination and duties in the quotation. - [RoboSense official Helios shop: exact H32F70 column ↗](https://store.robosense.ai/products/helios-series) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.robosense.ai/products/helios-series) 3D spinning LiDARCompare: RoboSense RS-Bpearl (offered 0.92 kg configuration) RoboSense ### RS-Bpearl (offered 0.92 kg configuration) Established comparison sensor for hemispherical near-field perception. 2,436 EURTax status is unclear on the reviewed product page; confirm the final price and shipping. Technically testable as a broad near-field sensor, but almost 1 kg sensor weight is unattractive for small drones. Mapping pipeline: Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Broad near-field coverage can support obstacle detection beside a façade. - Documentation and drivers have been available for some time. #### Weaknesses - Much heavier than JT16, JT128 or Airy. - Bpearl is sold in different revisions with different weights and accuracy specifications. IntegrationMedium Tie this price listing to its 0.92 kg specification; confirm exact revision, datasheet and package contents before ordering. Operating range / quantity For this documented configuration: 0.1–30 m at 10% reflectivity; manufacturer specifies a near blind zone ≤10 cm. Error / dispersion Associated manufacturer manual: typically ±3 cm on a 50% NIST target; do not transfer another Bpearl revision’s 1 cm precision. Interface 100 Mbit Ethernet/UDP; manufacturer SDK with ROS and ROS 2 supports RS-Bpearl. Specifications, availability & sources Rate / points 10/20 Hz in the associated older manual; approximately 576,000 points/s according to retailer. Weight / housing 0.92 kg without cable for this listing. · IP67 according to associated manual. Supply / power 13 W typical, 9–32 V DC in associated manufacturer documentation. Field of view / coverage 360° horizontal × 90° vertical, hemispherical. The manufacturer manual describes an older configuration; newer Bpearl documents give different values. Do not mix revisions into an idealized sensor. No proof for glass/water/spray or 50 cm ±5 cm nozzle distance. Availability when checked ROS Components offers backordering with an estimated 2–4 week lead time; immediate stock is not confirmed. Purchase / region European robotics specialist retailer; confirm international shipping and taxes in the quotation. - [RoboSense — Bpearl product page and near blind zone ↗](https://www.robosense.ai/en/rslidar/Bpearl) - [RoboSense — manual for the older 0.92 kg RS-Bpearl configuration, hosted by a specialist retailer ↗](https://www.roscomponents.com/wp-content/uploads/2024/11/Bpearl_User_Guide.pdf) - [ROS Components — Bpearl configuration, price and lead time ↗](https://www.roscomponents.com/product/rs-bpearl/) - [RoboSense — official SDK with ROS/ROS 2 support ↗](https://github.com/RoboSense-LiDAR/rslidar_sdk) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.roscomponents.com/product/rs-bpearl/) Directional hybrid 3D LiDARCompare: Seyond Robin E1X (192-line configuration) Seyond ### Robin E1X (192-line configuration) Low-power forward sensor for more distant façades and obstacles. 3,750 USDCRATUS sales terms: USD excluding taxes, duties, shipping, insurance and fees; other DDP quotations may differ. Unsuitable for direct measurement at a 50 cm sensor-to-surface distance because minimum range is 1 m. For a set-back sensor, calculate the actual sensor distance from the nozzle geometry. Mapping pipeline: Spatial point clouds become a map through registration, correctly timed IMU/pose data and SLAM. Georeferencing additionally requires suitable GNSS/INS data and calibration; the sensor alone does not control a cleaning flight. Mapping / SLAMPosition / navigationSurface / object recognition #### Strengths - Low power and a flat enclosure simplify mechanical integration. - A directional field of view can simplify processing a forward sector. #### Weaknesses - A 1 m minimum range excludes direct measurement at 50 cm; assess a larger sensor distance from the actual geometry. - T1 Ethernet requires a suitable host interface or media converter. IntegrationHigh Check Automotive Ethernet connection, supply and the exact line/scan variant. Provide a separate suitable near-range sensor. Operating range / quantity 1–250 m according to the current manufacturer page and 192-line brochure; 200 m at 10% reflectivity. Older 96-line documents specify up to 300 m. Error / dispersion ±5 cm ranging accuracy; 3 cm precision (1σ). Angular sampling of 0.1° × 0.1° is a different quantity. Interface Automotive Ethernet 1000BASE-T1, compatible with 100BASE-T1; UDP/TCP; PTP/gPTP/NTP. Official Seyond driver supports Robin E1X plus ROS and ROS 2. Specifications, availability & sources Rate / points 10–20 frames/s depending on configuration; 192 lines in the current datasheet. Weight / housing 470 g. · IP67 enclosure, IP69K window according to manufacturer. Supply / power 6 W according to associated manufacturer brochure; 9–16 V DC. Field of view / coverage 120° horizontal × 20° vertical; no surround view. 250/300 m and 96/192-line figures depend on revision/configuration. Manufacturer claims about mounting behind vehicle glass do not prove measurement of a glass façade. No demonstrated spray suitability or 50 cm ±5 cm guarantee. Availability when checked Publicly offered by CRATUS for USD 3,750; authorized US distributor US LiDAR also lists a purchase option. Confirm delivery for the 192-line variant. Purchase / region Manufacturer offices in the USA, Germany and China; specialist distribution documented in the USA and UK/Benelux. - [Seyond — current Robin E1X product page ↗](https://seyond.com/products/robin-e/) - [Seyond — 192-line brochure with accuracy, weight and power ↗](https://www.seyond.com/wp-content/uploads/2025/06/Robin-E1X_EN_automotive-1.pdf) - [Seyond — official ROS/ROS 2 driver ↗](https://github.com/Seyond-Inc/seyond_ros_driver) - [CRATUS — Robin E1X price ↗](https://www.cratustech.com/product/seyond-robin-e1x-mid-range-general-purpose-lidar-for-adas/) - [CRATUS — currency, tax and shipping terms ↗](https://www.cratustech.com/sales-terms-and-conditions/) - [US LiDAR — international purchasing channel ↗](https://us-lidar.com/seyond-lidar-sensors/) - [Digiflec — Seyond distribution in the UK and Benelux ↗](https://digiflec.com/seyond/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.cratustech.com/product/seyond-robin-e1x-mid-range-general-purpose-lidar-for-adas/) High-performance UAV surveying scannerCompare: RIEGL VUX-1UAV²² — scanner only RIEGL ### VUX-1UAV²² — scanner only Dense georeferenced corridor/terrain point clouds for professional surveying. Price on requestManufacturer quote depends on GNSS/INS, software and integration; no verified public fixed price. For separate surveying flights on a large platform; usually too heavy for compact façade-cleaning drones. Mapping pipeline: The line scanner produces a georeferenced 3D map only through flight motion and precise GNSS/IMU poses. Do not confuse a bare scanner with a complete VUX-SYS payload; not a cleaning-navigation system. Mapping / SLAM #### Strengths - High point rate and up to 15 returns in suitable PRR modes - Highly precise survey sensor with internal data storage #### Weaknesses - 3.5 kg sensor and 60 W before GNSS/INS/accessories - Near-range limitations and cooling complicate façade navigation IntegrationHigh Allow for a capable platform, GNSS/INS, boresight calibration, trajectory processing and RIEGL analysis software. Operating range / quantity PRR-dependent: at 50 kHz up to 755 m @20% / 1,415 m @80%; at 1,200 kHz 160 / 320 m; minimum distance 1.5–3 m Error / dispersion 10 mm measurement accuracy; 5 mm precision (1σ at 150 m) Interface Gigabit LAN, USB 2.0, GNSS RS232/PPS; 1 TB SSD Specifications, availability & sources Rate / points Up to 1.2 MHz PRR and 1.2 million measurements/s; 10–200 scans/s Weight / housing 3.5 kg without / 3.75 kg with fan; excluding external GNSS/IMU · IP64 Supply / power 11–34 V; typically 60 W excluding external GNSS/IMU Field of view / coverage Up to 360° line scan Maximum range and maximum point rate occur at different PRR settings. Ranging accuracy does not automatically equal georeferenced map accuracy. Availability when checked Current manufacturer datasheet dated 17 August 2026; specific system configuration on request. Purchase / region RIEGL Austria/Germany and international offices. - [RIEGL manufacturer product page, specifications and datasheet download ↗](https://www.riegl.com/en-usa/products/detail/riegl-vux-1uav-22) - [RIEGL VUX-1UAV²² datasheet, revision 17 August 2026 ↗](https://www.riegl.com/fileadmin/media/Products/04_UAV_based_Scanning/RIEGL_VUX-1UAV-22/RIEGL_VUX-1UAV-22_Datasheet_2026-08-17.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.riegl.com/en-usa/products/detail/riegl-vux-1uav-22) UAV surveying scannerCompare: RIEGL miniVUX-3UAV — scanner only RIEGL ### miniVUX-3UAV — scanner only Professional line scanner for georeferenced terrain/object point clouds. Price on requestConfiguration and price on manufacturer request; no publicly verified fixed unit price. Survey sensor for a separate mapping mission; payload mass and a 2 m minimum range fit close-range cleaning poorly. Mapping pipeline: A moving line scanner combined with a GNSS/IMU trajectory produces a georeferenced 3D point cloud. Distinguish the bare scanner from an integrated miniVUX-SYS/RiLOC payload; no autonomous cleaning navigation. Mapping / SLAM #### Strengths - Up to five returns per pulse and precise surveying data - Selectable scan/PRR modes for different mapping tasks #### Weaknesses - The bare scanner does not provide a complete surveyed trajectory - High mass, 2 m minimum distance and cooling requirements IntegrationHigh Add GNSS/INS, such as the optional RiLOC solution, a mount, boresight calibration, trajectory/RIEGL software and a data workflow. Operating range / quantity From 2 m; up to 170 m at ≥20% / 330 m at ≥80% reflectivity (mode-dependent) Error / dispersion 15 mm measurement accuracy; 10 mm precision under RIEGL test conditions Interface 2× Gigabit LAN; Wi-Fi; GNSS RS232/PPS; SD storage Specifications, availability & sources Rate / points PRR 100/200/300 kHz; up to 200,000 measurements/s; up to 100 scans/s Weight / housing 1.55 kg without / 1.60 kg with fan; excluding external GNSS/IMU · IP64 Supply / power 11–34 V; typically 18 W at 100 scans/s Field of view / coverage 360° at 100 kHz / reduced at 200 kHz; 180° at 200 kHz; 120° at 300 kHz Measurement accuracy does not equal complete map accuracy. 300 kHz PRR does not mean 300,000 output measurements/s. Not a substitute for obstacle/glass protection. Availability when checked Current RIEGL product with RiLOC system options; project-specific price and lead time. Purchase / region RIEGL Austria, Germany and worldwide office/partner network. - [RIEGL miniVUX-3UAV manufacturer page and system scope ↗](https://www.riegl.com/en-usa/products/detail/riegl-minivux-3uav) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.riegl.com/en-usa/products/detail/riegl-minivux-3uav) Integrated UAV surveying payloadCompare: YellowScan Mapper+ YellowScan ### Mapper+ Matched LiDAR/GNSS/IMU system for georeferenced 3D maps. Price on requestManufacturer quote depends on mount, software and accessories; no verified public fixed price. EUR/USD 0 shop placeholders are not prices. More complete survey payload for separate mapping missions; serves a substantially different purpose from real-time glass distance and cleaning navigation. Mapping pipeline: Integrated LiDAR/GNSS/IMU payload for a georeferenced 3D point cloud after trajectory and point-cloud processing. Includes more than a bare sensor, but does not replace autonomous cleaning or collision control. Mapping / SLAMPosition / navigation #### Strengths - LiDAR, GNSS/INS and boresight calibration as a matched system - Package includes GNSS antenna, batteries and POSPac UAV software #### Weaknesses - Higher mass and 35 W system power consumption - Clarify mapping software/licences and platform integration for each project IntegrationHigh Requires a suitable drone mount, GNSS correction/trajectory workflow and map processing; clarify CloudStation/LiveStation options separately. Operating range / quantity Recommended maximum flight altitude 100 m AGL; laser range up to 230 m Error / dispersion System accuracy 4 cm; system precision 3.5 cm according to manufacturer Interface Integrated Livox Avia + Applanix APX-15 UAV; Skyport/Gremsy or external power connection, USB data workflow Specifications, availability & sources Rate / points Up to 240,000 laser pulses/s, up to 3 returns; up to 10 Hz; manufacturer example 170 points/m² at 100 m / 10 m/s Weight / housing 1.1 kg without battery; manufacturer lists 1.3 kg for the supplied Mapper+ · No overall IP rating stated on the reviewed product page Supply / power 35 W; own battery or power through the platform interface Field of view / coverage 70.4° × 4.5° in the specified mapping configuration 100 m AGL is mission guidance; 230 m laser range is not a navigation clearance. System accuracy assumes suitable acquisition/GNSS conditions; no glass approval. Availability when checked Currently in the YellowScan range; request a binding package specification and delivery date. Purchase / region YellowScan France with a worldwide partner network; EU project support available. - [YellowScan Mapper+ manufacturer page, system specifications and package contents ↗](https://www.yellowscan.com/products/mapper-plus/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.yellowscan.com/products/mapper-plus/) Optical flow + ToFCompare: ARK Electronics ARK Flow (ARK-FLOW; standard, not MR/LR) ARK Electronics ### ARK Flow (ARK-FLOW; standard, not MR/LR) Relative ground motion to reduce drift; ToF height 250 USDExcludes taxes, duties and shipping under ARK sales terms; Ex Works Murray, Utah, unless otherwise agreed. Very lightweight supplementary measurement against ground-relative drift in suitable scenes. Position / navigation #### Strengths - Schematics, firmware and mechanical files are openly available. - PX4/ArduPilot integration and CAN firmware updates are documented. #### Weaknesses - Requires ground features, valid height and light; manufacturer specifies more than 9 lux, with an IR LED for low light. - No forward façade distance; test glossy ground and spray in the actual installation. IntegrationMedium Mount downwards, fuse flow and ToF, and define quality thresholds and spray protection. Operating range / quantity Flow from 0.08 m, optically without a fixed far-focus limit; integrated ToF, typically up to 30 m, limits the metric height reference Error / dispersion No general module accuracy for velocity or distance stated in the reviewed product overview Interface DroneCAN, two 4-pin JST-GH; PAW3902, AFBR-S50LV85D and IMU Specifications, availability & sources Rate / points Output rate not reliably specified in the reviewed excerpt Weight / housing 5 g; board 30 × 30 × 14 mm · Exposed module; no IP rating stated Supply / power 5 V; 71 mA average, 76 mA maximum Field of view / coverage ToF transmit beam 2° × 2°; flow FoV not stated on the reviewed standard product page Infinite optical working distance does not imply infinite metric navigation range; height, resolution and scene limit operation. Motion limit, not output rate: up to 7.4 rad/s flow motion; output timing not verified here. Availability when checked Manufacturer shop: in stock when checked. Purchase / region Manufacturer/shop offering from Utah, USA verified; destination countries are not guaranteed, so request delivery confirmation. - [ARK Flow standard: manufacturer specifications and price ↗](https://arkelectron.com/product/ark-flow/) - [ARK Flow official documentation ↗](https://docs.arkelectron.com/products/sensor/ark-flow) - [ARK Electronics prices and supply terms ↗](https://arkelectron.com/supply-terms-conditions/) - [PX4 flow concept and height reference ↗](https://docs.px4.io/main/en/sensor/optical_flow) - [ArduPilot: ground texture as the basis of flow ↗](https://software.ardupilot.org/plane/docs/common-optical-flow-sensors-landingpage.html) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://arkelectron.com/product/ark-flow/) Optical indoor positioning deckCompare: Bitcraze Lighthouse positioning deck (114991751) Bitcraze ### Lighthouse positioning deck (114991751) GNSS-free laboratory flight using optical base stations 95 USDUSD 95 for deck only; at least two Lighthouse V2 stations extra Useful for repeatable indoor development; manufacturer does not intend sunlit facade use. Position / navigation #### Strengths - Position is computed onboard without continuous pose radio transmission. - Lightweight introduction to optical local referencing. #### Weaknesses - Explicitly indoor-only; direct sunlight can degrade IR reception. - Other flight controllers need custom software; base stations are not included. IntegrationHigh Calibrate station geometry; develop a custom receiver/pose bridge for Cube/PX4. It is not a conventional rangefinder. Operating range / quantity Tracked volume limited by visibility and stations; manufacturer reports a 5 × 5 m test area Error / dispersion In a 5 × 5 m test: relative precision <1 mm, absolute accuracy <10 cm; these differ Interface UART / Crazyflie deck; observe maximum 3 V external signals Specifications, availability & sources Rate / points Depends on base stations, reception and firmware Weight / housing 2.7 g · Bare board; indoor system Supply / power Typically 30 mA; VCOM about 3.3–5.5 V Field of view / coverage Four IR receivers; stations above the drone and line of sight required Millimetre repeatability does not mean millimetre-accurate building coordinates. Not recommended for outdoor cleaning. Availability when checked Deck orderable from manufacturer store Purchase / region International store; check compatible base-station availability - [Bitcraze: Lighthouse Deck price ↗](https://store.bitcraze.io/products/lighthouse-positioning-deck) - [Bitcraze: Lighthouse system limits and accuracy ↗](https://www.bitcraze.io/documentation/system/positioning/ligthouse-positioning-system/) - [Bitcraze: Lighthouse hardware and external integration ↗](https://www.bitcraze.io/products/lighthouse-positioning-deck/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.bitcraze.io/products/lighthouse-positioning-deck) UWB positioning deckCompare: Bitcraze Loco positioning deck (114990655) Bitcraze ### Loco positioning deck (114990655) Small local test volume for GNSS-free navigation 95 USDUSD 95 for tag deck only; anchors extra, tax and shipping depend on destination Affordable learning and bench-testing option, primarily developed for indoor Crazyflie use. Position / navigation #### Strengths - Very small, light UWB tag. - Open hardware/software documentation and onboard position computation. #### Weaknesses - The tag cannot establish spatial coordinates without anchors. - No rugged outdoor cleaning integration; anchor nodes temporarily out of stock. IntegrationHigh Install at least four, preferably eight Loco Nodes; bridging to another flight controller requires custom development. Operating range / quantity 10 m tested in datasheet; no guaranteed outdoor coverage Error / dispersion ±10 cm UWB range per DWM1000 specification; position error depends on anchors Interface Crazyflie deck interface; no ready-made Cube/PX4 sensor Specifications, availability & sources Rate / points 500 ranging measurements/s shared across anchors; about 80 Hz per anchor with six anchors Weight / housing 3.3 g · Bare board; protect from water Supply / power Crazyflie deck supply; max. 150 mA Field of view / coverage UWB radio; antenna orientation matters Price does not include a complete system. Range error is not a guaranteed 3D position error. Availability when checked Deck orderable; Loco Nodes temporarily out of stock when checked Purchase / region Swedish manufacturer with international store - [Bitcraze: Loco Deck price ↗](https://store.bitcraze.io/collections/positioning/products/loco-positioning-deck) - [Bitcraze: Loco Deck datasheet ↗](https://www.bitcraze.io/documentation/hardware/loco_deck/loco_deck-datasheet.pdf) - [Bitcraze: positioning systems and anchor count ↗](https://www.bitcraze.io/documentation/system/positioning/) - [Bitcraze: anchor availability ↗](https://store.bitcraze.io/collections/positioning) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.bitcraze.io/collections/positioning/products/loco-positioning-deck) GNSS / RTKCompare: CubePilot Here4 Multiband RTK GNSS (HX4-06248) CubePilot ### Here4 Multiband RTK GNSS (HX4-06248) Global outdoor position; supplementary navigation 255 USDUS shop price; tax status not explicitly stated on the item. International import duties and VAT are paid by the customer under retailer terms. Lightweight GNSS for missions and a global reference; local distance sensors remain necessary. Position / navigation #### Strengths - Multiconstellation RTK and documented CAN integration for ArduPilot and PX4. - Integrated compass, IMU and barometer. #### Weaknesses - RTK requires a base or correction service; CubePilot specifies open-sky visibility and warns about built-up environments. - Building reflections and occlusion are critical near a façade; position does not replace surface measurement. IntegrationMedium Handle correction delivery and fix loss; check hardware/firmware versions, especially for moving baseline. Operating range / quantity Satellite position; no wall-distance measurement interval Error / dispersion RTK 0.01 m + 1 ppm CEP according to CubePilot; only with suitable reception and correction conditions Interface DroneCAN / CAN FD, NEO-F9P L1/L5; UART pins present Specifications, availability & sources Rate / points RTK up to 20 Hz according to manufacturer Weight / housing 60 g with cable · No IP rating stated in the reviewed documentation Supply / power 5 V; power consumption not stated in the reviewed manufacturer documentation Field of view / coverage Clear satellite visibility; no optical FoV specification No confirmed ±5 cm façade control. An integrated IMU alone does not establish enabled GNSS/INS odometry. Availability when checked IR-LOCK direct page: 160 units in stock/ready to ship when checked; stock can change. Purchase / region IR-LOCK US warehouse; international DHL/UPS options under its shipping policy, with destination and cost checked at checkout. - [CubePilot official Here4 manual ↗](https://github.com/CubePilot/cubepilot-docs/blob/master/here-4/here-4-manual.md) - [IR-LOCK Here4 standard: MPN, price and stock ↗](https://irlock.com/products/here4-multiband-rtk-gnss) - [IR-LOCK shipping from the US warehouse ↗](https://irlock.com/pages/shipping-1) - [IR-LOCK international duties ↗](https://irlock.com/pages/terms) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://irlock.com/products/here4-multiband-rtk-gnss) GNSS / RTKCompare: Holybro DroneCAN H-RTK F9P Helical (SKU 12037) Holybro ### DroneCAN H-RTK F9P Helical (SKU 12037) Global position and outdoor flight navigation; no direct wall distance 219 USDInternational DAP shipping from China; import tax, VAT and duties may be additional; shipping at checkout. Global navigation and a return point complement local façade sensors. Position / navigation #### Strengths - Dual-band ZED-F9P and a detachable helical antenna allow flexible antenna placement. - DroneCAN and ArduPilot/PX4 setup are documented. #### Weaknesses - RTK needs correction data from a base or service and suitable satellite geometry. - Occlusion and reflections near façades can degrade position and RTK fix; wall clearance is not measured. IntegrationMedium Check correction delivery, antenna position, CAN termination and behaviour during RTK float/fix loss. Operating range / quantity Satellite navigation; no near-distance measurement range Error / dispersion Manufacturer: 1.5 m with 3D fix; 0.01 m RTK under suitable conditions; no façade-clearance guarantee Interface DroneCAN 1 Mbit/s; UART2 for moving-baseline applications; USB for base operation Specifications, availability & sources Rate / points RTK up to 8 Hz; moving-base RTK up to 5 Hz; raw data up to 20 Hz Weight / housing 61.2 g according to comparison table · No IP rating for the helical variant in the reviewed listing Supply / power 4.75–5.25 V; approximately 250 mA Field of view / coverage Open sky required; no optical FoV specification Centimetre figures concern RTK positioning. They do not establish 50 cm clearance from glass. Availability when checked Direct shop: orderable/variant available. Shipping notice: office closed 22–26 October, processing afterwards. Purchase / region Holybro shipping policy states dispatch from China and international DAP shipping; confirm the destination at checkout. - [Holybro DroneCAN H-RTK F9P Helical: SKU, price and specifications ↗](https://holybro.com/products/dronecan-h-rtk-f9p-helical) - [Holybro RTK setup: base, rover and open sky ↗](https://docs.holybro.com/gps-and-rtk-system/zed-f9p-h-rtk-series/setup-and-getting-started-ardupilot) - [Holybro shipping, DAP and taxes ↗](https://holybro.com/policies/shipping-policy) - [CubePilot RTK antenna placement and difficult environments ↗](https://github.com/CubePilot/cubepilot-docs/blob/master/here-4/here-4-manual.md) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://holybro.com/products/dronecan-h-rtk-f9p-helical) Optical flow + ToFCompare: Holybro H-Flow Optical Flow / Distance (SKU 19006) Holybro ### H-Flow Optical Flow / Distance (SKU 19006) Relative horizontal velocity over the ground plus height measurement 189 USDInternational DAP shipping from China; import tax, VAT and duties may be additional; shipping at checkout. Supplementary drift measurement during GNSS outages while textured ground remains visible. Position / navigation #### Strengths - Flow, ToF and IMU combined in one lightweight module. - IR LED supports low light; manufacturer specifies more than 5 lux for flow. #### Weaknesses - Relative motion estimation needs visible ground texture, suitable lighting and valid height. - Test smooth, reflective or wet surfaces separately; downward flow does not measure frontal glass distance. IntegrationMedium Configure orientation, lever arm, flow scaling, quality and EKF fusion; perform calibration over suitable ground. Operating range / quantity Flow: stated 0.08–30 m; ToF typically up to 30 m, depending on target and environment Error / dispersion No general velocity or distance accuracy stated in the reviewed module overview Interface DroneCAN, 4-pin Pixhawk JST-GH; integrated ICM-42688-P IMU Specifications, availability & sources Rate / points Output rate not reliably specified in the reviewed excerpt Weight / housing 15.2 g with enclosure; 3.5 g without enclosure · No IP rating stated; design spray protection separately Supply / power Power through Pixhawk CAN connector; consumption not verified Field of view / coverage Flow up to 42°; ToF 12.4° × 6.2°, transmit beam 2° × 2° Optical flow is relative velocity information, not absolute façade localization or collision clearance. Motion limit, not output rate: flow motion up to 7.4 rad/s; this is not an image or data rate. Module output timing is not verified. Availability when checked Direct shop: variant available/orderable; shipping notice says processing resumes after the 22–26 October closure. Purchase / region Ships from China under an international DAP policy; confirm destination at checkout. - [Holybro H-Flow current price, SKU and technical overview ↗](https://holybro.com/products/h-flow) - [Holybro H-Flow installation and setup ↗](https://docs.holybro.com/peripherals/h-flow-dronecan/setup-guide.md) - [PX4 optical flow: distance measurement and fusion ↗](https://docs.px4.io/main/en/sensor/optical_flow) - [ArduPilot optical flow: ground texture and velocity measurement ↗](https://software.ardupilot.org/plane/docs/common-optical-flow-sensors-landingpage.html) - [Holybro shipping and duties ↗](https://holybro.com/policies/shipping-policy) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://holybro.com/products/h-flow) Active stereo/RGB-D camera with edge AICompare: Luxonis OAK-D Pro USB – IMX378 Luxonis ### OAK-D Pro USB – IMX378 Combines depth images, RGB documentation and local object detection. 429 USDOfficial US shop price for the base IMX378 variant; taxes/import charges and shipping depend on destination. OV9782 variant costs more. Useful when object detection or image analysis is required alongside spatial perception. Deliberately select a near-range profile for 50 cm. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - An IR dot projector and floodlight support depth with weak texture or low light. - RVC2 processes stereo and AI locally, reducing companion workload. #### Weaknesses - Standard 800P starts at about 0.8 m; 50 cm sensor-to-surface distance requires another configuration. - IMX378 RGB uses a rolling shutter; direct sunlight weakens the IR projector, and reflections/glass remain difficult. IntegrationMedium Run DepthAI and ROS 2 on a companion; match model conversion, camera/depth profiles and power supply. Deliberately choose fixed-focus/global-shutter RGB options for vibration. Operating range / quantity Ideal approximately 0.8–12 m; minimum depth approximately 0.8 m at 800P, 0.4 m at 400P or in extended disparity mode, 0.2 m at 400P plus extended mode. Error / dispersion Manufacturer median unit at 800P/75 mm baseline: <2% absolute depth error below 4 m, <4% at 4–7 m. Not a guaranteed maximum error for every unit. Interface USB-C, DepthAI/RVC2, integrated BNO086 IMU; official depthai-ros driver. Specifications, availability & sources Rate / points Depth rate depends on profile; up to 120 fps in suitable stereo configurations. 1280×800 is not generally guaranteed together with maximum filtering/AI load. Weight / housing Exact weight of this USB/RGB variant not unambiguously verified on the current primary pages reviewed here; add mount and cable. · IP66 only with the specified screw-lock USB cable; check the exact enclosure/cable revision. Supply / power Base streaming 2.5–3 W; AI, depth, encoder and IR add consumption. Current documentation recommends an external 15+ W supply/Y-adapter for Pro devices. Field of view / coverage Stereo 80° horizontal × 55° vertical; IMX378 RGB 66°×54°. IR floodlight does not produce night-time colour images: RGB does not see the IR. No verified measurement guarantee for spray, water films or 50 cm ±5 cm. Availability when checked Official shop displays ordering options and a delivery window; check stock for the exact focus/RGB variant. Purchase / region International manufacturer shop and regional distributors; taxes and shipping vary by destination. - [OAK-D Pro: current variant price ↗](https://shop.luxonis.com/products/oak-d-pro) - [OAK-D Pro: profiles, RGB sensor and power ↗](https://docs.luxonis.com/hardware/products/OAK-D%20Pro) - [Stereo accuracy under reference conditions ↗](https://docs.luxonis.com/hardware/platform/depth/depth-accuracy) - [IP66 requires the specified USB cable ↗](https://docs.luxonis.com/hardware/platform/environmental-specifications/ip-rating) - [Stereo and ToF limitations ↗](https://docs.luxonis.com/hardware/platform/depth/ToF-depth) - [Official ROS driver ↗](https://github.com/luxonis/depthai-ros) - [Stereo frame rate and pipeline cost ↗](https://docs.luxonis.com/software-v3/depthai/tutorials/optimizing) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://shop.luxonis.com/products/oak-d-pro) Wide-angle stereo/RGB-D camera with edge AICompare: Luxonis OAK-D Pro W USB – IMX378 Luxonis ### OAK-D Pro W USB – IMX378 Broad spatial perception and near-range coverage with fewer cameras. 529 USDOfficial US shop price for the IMX378 version; OV9782 version costs more. Taxes, import charges and shipping depend on destination. Useful for broad collision/environment perception and a nominal 50 cm near range; lower angular pixel density than standard OAK-D Pro. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - The large field of view reduces lateral blind areas and can replace several narrower cameras. - Active IR stereo, IR floodlight and local AI combined in a lightweight enclosure. #### Weaknesses - Wide angle spreads pixels over a larger angle and reduces accuracy, especially at the edges. - IMX378 RGB uses a rolling shutter; direct sunlight, glass and wet reflections remain limitations despite IR. IntegrationMedium Connect USB to a companion and use DepthAI/ROS 2; account for wide-angle calibration, RGB-depth overlap and edge masks. Send only validated derived data to the flight controller. Operating range / quantity Ideal 0.4–6 m; minimum depth approximately 0.37 m at 800P or 0.2 m with an extended profile. Advertised 12 m is maximum effective range at the image centre, not precision working range. Error / dispersion Manufacturer median unit at 800P: <2% absolute depth error below 3.5 m, <4% at 3.5–6.5 m; worse towards the image edges. Interface USB-C 3.2 Gen 1, RVC2/DepthAI, BNO086 IMU and depthai-ros. Specifications, availability & sources Rate / points Sensor pair up to 1280×800; actual depth rate depends on profile, filters and AI. Maximum sensor rate is not a depth-pipeline guarantee. Weight / housing 91 g; 97×29.5×23.1 mm, without cable/mount. · IP66 only with the specified screw-lock USB cable. Supply / power Product page: up to 7.5 W; current Pro documentation recommends an external 15+ W supply/Y-adapter. Illumination and pipeline affect consumption. Field of view / coverage Stereo 127° horizontal × 79.5° vertical, 150° diagonal; IMX378 RGB 95°×72°. 150° is the diagonal stereo field of view, not the horizontal value. IP66 and near-range minimum depth do not establish ±5 cm control on a glass façade during spray. Availability when checked Shop offers IMX378/OV9782 variants; dynamic stock messages conflict, so confirm availability before ordering. Purchase / region International Luxonis shop and distributors; check shipping to Germany and import costs at checkout. - [OAK-D Pro W: current IMX378 price ↗](https://shop.luxonis.com/products/oak-d-pro-w) - [OAK-D Pro W: current technical product details ↗](https://checkout.luxonis.com/products/oak-d-pro-w) - [Manufacturer accuracy measurements ↗](https://docs.luxonis.com/hardware/platform/depth/depth-accuracy) - [Specified USB cable required for IP66 ↗](https://docs.luxonis.com/hardware/platform/environmental-specifications/ip-rating) - [Current Pro power guidance ↗](https://docs.luxonis.com/hardware/products/OAK-D%20Pro) - [Stereo/ToF limitations ↗](https://docs.luxonis.com/hardware/platform/depth/ToF-depth) - [Official ROS driver ↗](https://github.com/luxonis/depthai-ros) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://shop.luxonis.com/products/oak-d-pro-w) Ultrasonic localization systemCompare: Marvelmind Starter Set Industrial Super-MP-3D Marvelmind ### Starter Set Industrial Super-MP-3D Local 3D position from fixed ultrasonic beacons 1,199 EUREUR 1199 for five Industrial Super-Beacons and Super-Modem; check tax/shipping at checkout A GNSS-free experiment using local infrastructure; it does not measure wall distance. Position / navigation #### Strengths - The kit can provide four stationary beacons and one moving beacon. - Ultrasound adds a measurement principle independent of camera/LiDAR. #### Weaknesses - Rotors, water jets, wind and acoustic occlusion need practical testing. - Local beacons and suitable geometry required; range does not establish coverage. IntegrationHigh Secure beacons, verify the local map, mount the mobile beacon with vibration control and fuse position with IMU; validate a version-specific ExternalNav bridge. Operating range / quantity Recommended acoustic path up to 30 m depending on frequency and environment Error / dispersion Super-MP family: typical ±2 cm in suitable conditions; unproven in rotor wash/spray Interface RS485 / UART / USB; Super-Modem also Wi-Fi UDP Specifications, availability & sources Rate / points Depends on architecture, beacon count and fusion; no universal flight-control rate Weight / housing Set mass not specified; confirm mobile beacon mass with manufacturer · Industrial beacons up to IP56; some connectors IP67 Supply / power External 5 V; provide 1 A peak supply; no battery by default Field of view / coverage Clear acoustic line of sight to sufficient beacons required This acoustic system localizes between beacons; it does not replace a wall/glass ultrasonic rangefinder. Industrial/outdoor hardware is described, but high-pressure cleaning performance is not validated. Availability when checked Industrial kit listed for ordering by manufacturer Purchase / region Estonian international supplier; select matching 868/915 MHz radio variant - [Marvelmind: Industrial Set product and price ↗](https://marvelmind.com/product/starter-set-industrial-super-mp-3d/) - [Marvelmind: Super-MP operating principle and typical accuracy ↗](https://marvelmind.com/product/starter-set-super-mp-3d/) - [Marvelmind: deployment and line of sight ↗](https://marvelmind.com/pics/Marvelmind_Robotics_ENG_placement_manual.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://marvelmind.com/product/starter-set-industrial-super-mp-3d/) VIO computer with IMU and flight controlCompare: ModalAI VOXL 2 Board Only (MDK-M0154-1-00) ModalAI ### VOXL 2 Board Only (MDK-M0154-1-00) Onboard computing platform for visual local navigation 1,399.99 USDUSD 1399.99 for board only on opened product page; cameras, supply, cooling and protection extra Suitable platform for VIO with fixed marker references; board price is not a complete positioning sensor. Position / navigation #### Strengths - Compact platform with IMUs and a VIO software ecosystem. - Manufacturer documents local pose and AprilTag relocalization. #### Weaknesses - Requires cameras and accurate camera/IMU calibration. - Glass, mist, contamination and featureless surfaces degrade visual tracking. IntegrationHigh Add a compatible tracking camera; align intrinsics/extrinsics, timestamps and firmware versions; handle quality degradation and reset jumps. Operating range / quantity No intrinsic ranging interval; depends on separately selected camera and scene Error / dispersion VIO accumulates drift; no guaranteed absolute centimetre accuracy Interface MIPI cameras; PX4/ArduPilot integration, MAVLink / ROS via VOXL SDK Specifications, availability & sources Rate / points Depends on cameras, VIO configuration and compute load Weight / housing 16 g board excluding cameras/accessories · Exposed electronics; no spray-environment IP rating specified Supply / power Manufacturer lists 0.1–7 W for board; cameras/add-ons extra Field of view / coverage Depends on separately purchased tracking/stereo cameras The older VOXL 2 Flight Deck is officially EOL and is not recommended as a current kit here. The VOXL 2 board alone cannot observe the environment or provide visual position. Availability when checked Active production; product page states shipment within 30 business days Purchase / region US manufacturer supplying internationally; confirm import/delivery terms - [ModalAI: VOXL 2 board product, exact SKU and price ↗](https://www.modalai.com/products/voxl-2?variant=39914779836467) - [ModalAI: VIO setup and limitations ↗](https://docs.modalai.com/flying-with-vio/) - [ModalAI: AprilTag relocalisation ↗](https://docs.modalai.com/voxl-vision-hub-apriltag-relocalization/) - [ModalAI: end-of-life list for the earlier Flight Deck ↗](https://www.modalai.com/pages/eol) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.modalai.com/products/voxl-2?variant=39914779836467) UWB tag / anchorCompare: Nooploop LinkTrack P-B (LTP-B) Nooploop ### LinkTrack P-B (LTP-B) Local radio coordinate system without GNSS Price on requestRequest a quote; specify anchor count, antennas and regional radio variant A local position aid with several surveyed anchors; a dedicated nozzle range sensor is still required. Position / navigation #### Strengths - The tag computes local coordinates without satellites or RTK. - The same hardware can serve as a tag or anchor. #### Weaknesses - Typical 3D error substantially exceeds the ±5 cm nozzle target. - Metal, occlusion and poor anchor geometry can bias measurements. IntegrationHigh Calibrate anchors at different elevations; fuse timestamped tag measurements with IMU data and supply External Navigation. Operating range / quantity Selection guide V1.3: recommended P-B radio distance 300 m; facade performance unproven Error / dispersion LinkTrack: typical 10 cm in 1D/2D, 30 cm in 3D; system figures, no ±5 cm guarantee Interface UART / USB; ROS protocol integration Specifications, availability & sources Rate / points Up to 200 Hz depending on system configuration Weight / housing 34.3 g excluding external antenna, selection guide V1.3 · IP66 in selection guide V1.3; assess connectors separately Supply / power 3.6–5.5 V; 1.35 W, selection guide V1.3 Field of view / coverage Radio coverage depends on antenna and installation Exact P-B hardware data comes from selection guide V1.3; longer distances on the current series page partly concern other variants. Native PX4/Cube compatibility is not established. Availability when checked Manufacturer lists the P-B series; confirm current revision and radio approval Purchase / region International manufacturer supply; check regional radio variant - [Nooploop: LinkTrack product and system data ↗](https://www.nooploop.com/linktrack/) - [Nooploop: exact variants, selection guide V1.3 ↗](https://ftp.nooploop.com/downloads/linktrack/LinkTrack_and_Parts_Selection_Guide_Manual_V1.3_zh.pdf) - [Nooploop: integration and protocols ↗](https://support.nooploop.com/linktrack/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.nooploop.com/linktrack/) External motion-capture cameraCompare: OptiTrack PrimeX 13 OptiTrack ### PrimeX 13 Reference measurement and local flight position using multiple external cameras 2,499 USDFrom USD 2499 per camera per manufacturer; multiple cameras, Motive licence, markers and networking extra Strong reference for a controlled test volume; substantial setup for a complete outdoor facade. Position / navigation #### Strengths - The drone carries only a lightweight marker arrangement. - Independent reference for measuring estimator drift and control error. #### Weaknesses - Drone structure, nozzle boom and hose can occlude tracking. - A single-camera price is not a complete system budget. IntegrationHigh Calibrate multiple cameras and a rigid marker body; transform coordinates/timebase and monitor pose quality and link failures. Operating range / quantity Capture volume depends on camera count, lenses, marker size and installation Error / dispersion System/calibration-dependent; one camera alone does not provide advertised system accuracy Interface Ethernet; Motive / NatNet, custom MAVLink pose bridge Specifications, availability & sources Rate / points Up to 240 frames/s in manufacturer robotics brochure Weight / housing Stationary cameras remain on the ground; only markers are airborne · Plan weather/spray protection for the installation Supply / power Stationary Ethernet/PoE system; plan network power separately Field of view / coverage Lens-dependent; overlapping fields of view required Validate in a controlled volume. No guarantee for outdoor light, rain or arbitrary facade size; not a ready-made autonomous flight controller. Availability when checked PrimeX 13 listed; plan the system through configurator/sales Purchase / region International manufacturer/partner network - [OptiTrack: PrimeX 13 product ↗](https://www.optitrack.com/cameras/primex-13) - [OptiTrack: robotics brochure with camera prices and frame rates ↗](https://optitrack.com/api/media/file/Robotics-Brochure-2025.pdf) - [PX4: external motion-capture position ↗](https://docs.px4.io/main/en/ros/external_position_estimation) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.optitrack.com/cameras/primex-13) Area iToF/RGB-D cameraCompare: Orbbec Femto Bolt Orbbec ### Femto Bolt Alternative depth method for laboratory work, indoor mapping and material tests. 418 USDOfficial manufacturer shop: taxes and shipping calculated at checkout; not a German VAT-inclusive price. Useful as an iToF comparison sensor and in protected indoor spaces; poorly suited as an exposed outdoor façade sensor on the cleaning head. Mapping / SLAMSurface / object recognition #### Strengths - Area time-of-flight sensing does not need stereo correspondence from visible texture. - Selectable wide/normal-angle profiles, RGB and IMU support laboratory RGB-D and material comparisons. #### Weaknesses - Tight specified temperature/humidity limits, no verified IP rating and 348 g argue against unprotected spray use. - iToF remains sensitive to multipath, glass, reflections and mutual IR interference; prove outdoor operation separately. IntegrationMedium Connect USB to a companion, use Orbbec SDK/OrbbecSDK_ROS2, and select a suitable depth mode and adequate power. Depth processing requires a host; not an autonomous flight-controller sensor. Operating range / quantity WFOV: 0.25–2.21 m unbinned or 0.25–2.88 m binned. NFOV: 0.5–3.86 m unbinned or up to 5.46 m with 2×2 binning; maxima are not all simultaneous. Error / dispersion Manufacturer: typical systematic error <11 mm +0.1% of distance, random standard deviation ≤17 mm under specified conditions. No guarantee for wet glass façades. Interface USB 3.2 Gen 1 Type-C; 12 V DC or USB-C power, 8-pin synchronization, Orbbec SDK and ROS 2 wrapper. Specifications, availability & sources Rate / points WFOV 1024×1024 up to 15 fps; NFOV 640×576 up to 30 fps; RGB up to 3840×2160/30 fps on the current product page. Weight / housing 348 g according to current hardware documentation; older datasheets state 335 g. · No verified IP rating; manufacturer operating environment 10–25 °C and 8–90% relative humidity, non-condensing. Supply / power 4.35 W average; up to 8.7 W with DC supply or 7.7 W through Type-C according to hardware documentation. Field of view / coverage Depth WFOV 120°×120° or NFOV 75°×65°; RGB 80°×51°. iToF provides metric depth, not thermal imaging. The 0.25 m and 5.46 m limits come from different modes; 50 cm ±5 cm in spray is not confirmed. Availability when checked Official shop showed out of stock/sold out when checked; confirm actual retailer stock. Purchase / region International Orbbec shop, EU/US power adapters and distributors; delivery timing unresolved. - [Femto Bolt: current price and stock shortage ↗](https://store.orbbec.com/products/femto-bolt) - [Femto Bolt: technical data and accuracy definitions ↗](https://www.orbbec.com/products/tof-camera/femto-bolt/) - [Current hardware specification: weight, operating environment and peak power ↗](https://doc.orbbec.com/documentation/Orbbec%20Femto%20Bolt%20Documentation/Femto%20Bolt%20Hardware%20Specifications) - [Manufacturer datasheet: depth modes and interfaces ↗](https://www.orbbec.com/wp-content/uploads/2023/08/ORBBEC_Datasheet_Femto-Bolt-3.pdf) - [All four depth modes ↗](https://www.orbbec.com/staging/wp-content/uploads/2023/08/ORBBEC_Product-Brief_Femto-Bolt_0713.pdf) - [Official Orbbec ROS 2 wrapper ↗](https://github.com/orbbec/OrbbecSDK_ROS2) - [ToF: surfaces, multipath and interference ↗](https://docs.luxonis.com/hardware/platform/depth/ToF-depth) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.orbbec.com/products/femto-bolt) Active/passive stereo/RGB-D cameraCompare: Orbbec Gemini 335L Orbbec ### Gemini 335L Lightweight USB option for façade RGB-D and point clouds with extended near range. 359 USDOfficial manufacturer shop: taxes and shipping calculated at checkout; not a German VAT-inclusive price. An interesting trial option around 50 cm and for RGB-D mapping; nominal near range does not prove the ±5 cm requirement. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - Global shutter for depth and RGB plus hardware synchronization help with moving captures. - 95 mm stereo baseline, internal depth processing and ROS 2 support reduce companion workload compared with processing raw stereo images. #### Weaknesses - 20 m+ is a scene-dependent upper limit, not a useful precision working range for every façade. - Manufacturer marketing about sunlight and translucent materials does not replace testing on reflective or wet glass. IntegrationMedium Connect USB to a companion, use OrbbecSDK_ROS2 and a calibrated camera pose; evaluate point-cloud/depth quality before feeding the autopilot interface. Operating range / quantity Ideal 0.25–6 m; nominal operating range 0.17–20 m+ depends on profile and scene, not guaranteed façade range. Error / dispersion Manufacturer calibration metric: ≤0.8% at 2 m (1280×800, 90×90% ROI), ≤1.6% at 4 m (80×80% ROI). Not maximum error for every surface. Interface USB 3.0/2.0 Type-C, UVC, Orbbec SDK v2; 8-pin synchronization and 6-DoF IMU. Specifications, availability & sources Rate / points Depth up to 1280×800/30 fps or 848×530/60 fps; RGB up to 1280×800/60 fps. Weight / housing 133 g ±3 g; 124×29×27 mm. · IP65 with suitable cables; do not confuse this with the compact Gemini 335’s IP5X. Supply / power 5 V USB; average <3 W, peak <6 W according to product table. Field of view / coverage Depth 90°×65° ±3° at 2 m; RGB 94°×68° ±3°. IP protection requires suitable cables. No confirmed suitability for spray, a fogged window or cleaning agents; calibration metrics do not automatically apply at 0.5 m. Availability when checked Official shop: in stock and ready for shipping on the check date. Purchase / region International manufacturer shop lists Germany as a destination; distributor network in Europe, North America and Asia. - [Gemini 335L: current price and product data ↗](https://store.orbbec.com/products/gemini-335l) - [Gemini 330: protection, accuracy conditions and cables ↗](https://www.orbbec.com/gemini-330-series/) - [Gemini 330 datasheet: ROI conditions ↗](https://orbbec.com/wp-content/uploads/2024/04/Gemini-330-series-Datasheet-V1.4_20250124.pdf) - [Official Orbbec ROS 2 wrapper ↗](https://github.com/orbbec/OrbbecSDK_ROS2) - [General stereo surface limitations ↗](https://docs.luxonis.com/hardware/platform/depth/ToF-depth) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.orbbec.com/products/gemini-335l) Active stereo/RGB-D cameraCompare: RealSense Depth Camera D455 RealSense ### Depth Camera D455 Affordable USB entry point for spatial perception and RGB-D mapping. 499 USDOfficial US shop; local taxes, duties and shipping extra. Shop mentions a tariff surcharge since 3 February 2026; check the final amount. Useful for initial RGB-D/point-cloud experiments with a companion computer; 50 cm is already below the stated minimum depth at full resolution. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - Global shutter on depth and RGB sensors reduces motion distortion. - Integrated IMU and a widely used ROS 2 driver simplify time-aligned RGB-D mapping. #### Weaknesses - The standard enclosure and USB connection need integration suitable for the spray area. - Glass and glossy wet façades can produce false depth or holes; the IR projector cannot supply useful texture in every outdoor scene. IntegrationMedium Connect USB to a Linux companion, use the SDK and realsense-ros; configure drone extrinsics, timestamps and quality filtering. No direct Pixhawk distance-sensor interface. Operating range / quantity Ideal 0.6–6 m; minimum depth about 0.52 m at maximum resolution. Error / dispersion Manufacturer: <2% at 4 m. This does not establish accuracy at 0.5 m. Interface USB-C 3.1 Gen 1; RealSense SDK/librealsense and ROS 2 wrapper. Specifications, availability & sources Rate / points Depth up to 1280×720; up to 90 fps depending on profile. RGB 1280×800 at 30 fps; do not assume all maxima form one profile. Weight / housing 116 g nominal, without cable/mount. · No demonstrated IP protection for this standard variant. Supply / power Powered through USB 3.1; consumption for the chosen combined stream/projector profile is not verified here. Field of view / coverage Depth 87°×58°; RGB 90°×65° (horizontal × vertical). No reliable clearance guarantee is established for spray, fogged optics or reflective glass. Mapping and 50 cm control require separate suitability validation. Availability when checked Listed as in stock in the official shop when checked. Purchase / region Manufacturer shop lists Germany and numerous countries in Europe, the Americas and Asia; check import charges. - [D455: current shop price and specifications ↗](https://store.realsenseai.com/buy-realsense-depth-camera-d455.html) - [D400 family: D455 mechanical data ↗](https://www.intelrealsense.com/wp-content/uploads/2024/10/Intel-RealSense-D400-Series-Datasheet-October-2024.pdf) - [Optical filters and reflection errors ↗](https://dev.realsenseai.com/docs/optical-filters-for-intel-realsense-depth-cameras-d400/) - [Official ROS 2 wrapper ↗](https://github.com/realsenseai/realsense-ros) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.realsenseai.com/buy-realsense-depth-camera-d455.html) Rugged active stereo/RGB-D cameraCompare: RealSense Depth Camera D457 GMSL/FAKRA RealSense ### Depth Camera D457 GMSL/FAKRA Mechanically locked camera connection for an integrated Jetson system. 599 USDOfficial US shop; local taxes, duties and shipping extra. Displayed price is not a German VAT-inclusive price; check tariff surcharges at checkout. Suitable for a planned Jetson/GMSL integration; at full resolution it has the same near-range conflict around 50 cm as D455. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - A locking FAKRA connector and IP65 provide a more robust drone installation than an unsecured USB connection. - Global shutter, IMU and hardware synchronization support moving RGB-D captures. #### Weaknesses - Deserializer, suitable carrier board and drivers increase cost, weight and integration effort. - A rugged enclosure does not improve stereo measurement through mist, droplets or reflective façades. IntegrationHigh Choose a GMSL deserializer with drivers matching the JetPack/kernel version, then run librealsense/ROS 2 on the companion. MAVLink output requires custom processing and integration. Operating range / quantity Ideal 0.6–6 m; minimum depth about 0.52 m at maximum resolution. Error / dispersion Manufacturer: <2% at 4 m under reference conditions; no wet-glass guarantee. Interface GMSL over FAKRA; Maxim deserializer required at the host; not a simple USB camera replacement. Specifications, availability & sources Rate / points Depth up to 1280×720 and up to 90 fps depending on profile; RGB 1280×800 at 30 fps. Weight / housing Approximately 145 g, excluding cable and deserializer. · IP65 according to manufacturer; entire assembly and suitable cables must maintain sealing. Supply / power Power through GMSL/FAKRA; consumption depends on profile. Budget separately for deserializer/carrier board and companion. Field of view / coverage Depth 87°×58°; RGB 90°×65°. IP65 does not demonstrate correct measurement under a direct cleaning jet or chemical exposure. No ±5 cm guarantee at 50 cm. Availability when checked Official shop: in stock on the check date. Purchase / region International manufacturer shop including Germany; check regional cable/carrier availability separately. - [D457: current shop price, stock and specifications ↗](https://store.realsenseai.com/buy-realsense-depth-camera-d457.html) - [D457 datasheet, August 2025 revision ↗](https://realsenseai.com/wp-content/uploads/2025/08/RealSense-D457-Datasheet-Aug_2025.pdf) - [GMSL/FAKRA integration ↗](https://www.realsenseai.com/products/d457-gmsl-fakra/) - [Stereo reflection limitations ↗](https://dev.realsenseai.com/docs/optical-filters-for-intel-realsense-depth-cameras-d400/) - [RealSense ROS 2 ↗](https://github.com/realsenseai/realsense-ros) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.realsenseai.com/buy-realsense-depth-camera-d457.html) Active stereo/RGB-D camera with EthernetCompare: RealSense Depth Camera D555 PoE RealSense ### Depth Camera D555 PoE Image/depth acquisition over Gigabit Ethernet and PoE. 649 USDOfficial US shop; local taxes, shipping and duties extra. Check the final amount/tariff surcharge at checkout. Interesting for an Ethernet topology; significantly heavier than USB stereo. A suitable reduced-depth profile is required for 50 cm. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - One Ethernet cable can carry both data and power. - SDK/DDS enables platform-independent companion integration; global shutter and IMU support mobile perception. #### Weaknesses - 337 g plus PoE supply is a substantial addition for small drones. - Conflicting manufacturer power/peak-frame-rate figures and current lack of stock complicate immediate selection. IntegrationHigh Configure PoE converter/PSE, Gigabit port and MTU/DDS network; verify ROS 2/SDK/firmware compatibility for the exact release. Not a direct autopilot rangefinder. Operating range / quantity Ideal 0.6–6 m; minimum depth about 0.52 m at maximum resolution, 0.26 m in VGA mode. Error / dispersion No independent current D555 accuracy metric verified in the reviewed datasheet; do not transfer D455 values wholesale. Interface Gigabit Ethernet/RJ45, PoE and DDS in RealSense SDK; USB-C for power/synchronization and production/debug use. Specifications, availability & sources Rate / points Datasheet v1.1: 1280×720 at up to 30 fps over PoE, smaller profiles up to 60 fps. Shop advertises up to 90 fps: establish the exact firmware profile before purchasing/implementing. Weight / housing 337 g nominal. · IP65; enclosure temperature −20 to +50 °C according to manufacturer. Supply / power Manufacturer figures conflict: shop 5.5 W typical, datasheet v1.1 3.5 W. Datasheet requires a PoE PSE ≥15 W; size supply and converter with headroom. Field of view / coverage Depth 87°×58°; RGB 90°×65°. PoE is an interface, not an additional measurement method. Spray, glass and water films remain optical limitations; IP65 alone does not guarantee depth quality. Availability when checked Official shop showed out of stock/backorder when checked; delivery date not verified. Purchase / region Official shop lists Germany and many international destinations; confirm actual delivery availability separately. - [D555: current price, stock shortage and shop claims ↗](https://store.realsenseai.com/buy-realsense-depth-camera-d555.html) - [D555 datasheet v1.1: profiles, weight, power and DDS ↗](https://dev.realsenseai.com/download/42013/) - [RealSense ROS 2 integration ↗](https://github.com/realsenseai/realsense-ros) - [Optical stereo limitations ↗](https://dev.realsenseai.com/docs/optical-filters-for-intel-realsense-depth-cameras-d400/) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://store.realsenseai.com/buy-realsense-depth-camera-d555.html) Passive stereo camera with externally computed depthCompare: Stereolabs ZED X Mini — 2.2 mm wide angle Stereolabs ### ZED X Mini — 2.2 mm wide angle RGB stereo, visual odometry and mapping on Jetson. 549 USDOfficial camera price; tax status not unambiguously stated. Jetson, GMSL capture card, cable and shipping extra. Useful for mapping, VIO and close 3D on textured surfaces; requires a more powerful companion system. Mapping / SLAMSurface / object recognitionInspection / documentation #### Strengths - Global shutter and a 50 mm stereo baseline provide broad near-range perception without an active IR projector. - ZED SDK and ROS 2 wrapper provide depth, point clouds, odometry and mapping functions. #### Weaknesses - The camera does not compute depth maps alone; Jetson, capture card and their energy/weight belong in the system cost. - Low light, low-contrast and reflective surfaces complicate correspondence; an optional polarizer does not generally solve transparent glass. IntegrationHigh Match compatible Jetson/carrier, ZED Link, drivers, ZED SDK and ROS 2. No direct USB operation or Pixhawk interface. Operating range / quantity 2.2 mm variant: ideal 0.1–4 m; maximum hardware range 0.1–8 m. Do not transfer values from the 4 mm/narrow variant. Error / dispersion Manufacturer: <1.6% at 2 m, <3% at 4 m for the 2.2 mm variant; no guaranteed error at 50 cm on glass. Interface GMSL2 through FAKRA-Z, ZED Link capture card/ZED Box and a supported NVIDIA Jetson; ZED SDK/ROS 2. Specifications, availability & sources Rate / points 2×1920×1200 at 60 fps; reduced profiles up to 120 fps. Depth rate also depends on SDK/Jetson profile. Weight / housing 151 g; 93.6×31.8×36.7 mm. · Current technical specification lists IP67 and −20 to +60 °C; check certification for the exact ordering SKU. Supply / power Camera typically 1.46 W over GMSL2/PoC; Jetson and capture card extra. Field of view / coverage 2.2 mm optics: 110° horizontal × 80° vertical, 120° diagonal. IP67 describes enclosure protection, not correct depth measurement in spray. Current manufacturer pages partly differ from older IP66 specifications. Availability when checked Manufacturer states ZED X family shipping within 2–3 days; confirm the exact Mini/optics variant when ordering. Purchase / region Manufacturer in France with international shop/distributors; confirm destination and optics variant. - [ZED X family: current prices and shipping notice ↗](https://www.stereolabs.com/products/zed-x) - [ZED X/Mini: current technical specifications and polarizer limits ↗](https://docs.stereolabs.com/docs/products/cameras/zedx/specifications) - [Jetson and GMSL2 capture card ↗](https://support.stereolabs.com/articles/2878711998-how-can-i-install-the-zed-x-on-nvidia-jetson-development-kits) - [No USB converter connection for ZED X ↗](https://support.stereolabs.com/articles/2126344525-can-i-connect-the-zed-x-and-zed-x-mini-to-a-usb3-port) - [ZED ROS 2 ↗](https://docs.stereolabs.com/docs/integrations/ros-2) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.stereolabs.com/products/zed-x) mmWave radar development boardCompare: Texas Instruments IWR6843AOPEVM (development board) Texas Instruments ### IWR6843AOPEVM (development board) Radar research: point cloud and custom distance/velocity estimation 206.51 USDDigiKey US unit price; tax status not explicitly stated on the product page. US tariffs may be added in the basket. Technology experiment for a future custom radar system; substantial development effort. Surface / object recognition #### Strengths - TI provides an SDK, reference designs and hardware documentation. - Broad field of view and radar data enable custom processing. #### Weaknesses - The EVM is not a finished UAV rangefinder; signal processing, target selection and autopilot integration require application-specific development. - No demonstrated window detection or flight-control accuracy; enclosure and thermal design are required. IntegrationHigh Develop chirps, DSP, tracking, outlier handling and a message adapter; check the board revision. Operating range / quantity No fixed rangefinder measurement interval; depends on chirp profile, target and algorithm Error / dispersion No general distance accuracy guaranteed for the complete board; must be developed and measured for the application Interface USB for point clouds; expansion connectors, optional MMWAVEICBOOST / DCA1000 Specifications, availability & sources Rate / points Profile/firmware dependent; no ready-made distance output rate Weight / housing Not stated in the reviewed product overview · Exposed EVM; no verified IP rating Supply / power USB power; consumption and additional development adapters depend on the setup Field of view / coverage 120° horizontal × 120° vertical; 3 TX / 4 RX; 60–64 GHz A radar point cloud is not automatically metric odometry or a validated clearance safeguard. Availability when checked DigiKey: zero stock; estimated arrival 14 January 2027, not guaranteed. TI stock display is unclear without signing in. Purchase / region US DigiKey listing verified; check delivery and 60–64 GHz regulatory use for each destination, without assuming worldwide supply. - [TI IWR6843AOPEVM overview and development tools ↗](https://www.ti.com/tool/IWR6843AOPEVM) - [DigiKey: exact EVM variant, current price and stock ↗](https://www.digikey.com/en/products/detail/texas-instruments/IWR6843AOPEVM/12165115) - [TI EVM user guide, revision E / variant notes ↗](https://www.ti.com/lit/ug/swru546e/swru546e.pdf) Checked: 22.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.digikey.com/en/products/detail/texas-instruments/IWR6843AOPEVM/12165115) Liquid / cleaning-line pressureCompare: ifm PT5400 (0–400 bar) ifm ### PT5400 (0–400 bar) Pressure monitoring at the pump or in an appropriately pressure-rated cleaning line 176.10 USDDigiKey US unit price; tax and shipping excluded or determined at checkout. Useful for pressure-drop and pump monitoring; does not measure wall distance. Monitor cleaning process #### Strengths - Compact industrial housing with a fast analogue output. - A 400-bar measuring range suits high-pressure experiments better than a 10-bar hobby sensor. #### Weaknesses - Pressure at the ground pump does not automatically equal nozzle pressure after a long hose. - Fittings, pulsation, chemical compatibility and mechanical pressure protection must match the complete system. IntegrationMedium Provide a current-loop receiver/ADC or I/O gateway; detect pressure spikes and line failure. No direct PX4/Cube plug-in connection. Operating range / quantity 0–400 bar gauge pressure; static pressure rating 1,000 bar, minimum burst pressure 1,700 bar. These limits do not extend the measuring range. Error / dispersion Characteristic deviation <±0.5% of span, i.e. <±2 bar at 400 bar full scale; allow separately for temperature effects. ±0.05% describes repeatability only. Interface 4–20 mA, M12; G1/4 male process thread Specifications, availability & sources Rate / points Analogue step response 1 ms; reporting rate depends on the ADC and software bridge Weight / housing 60 g excluding application fittings and cable · IP67 / IP69K; wetted 17-4 PH stainless steel, FKM process seal Supply / power 8.5–36 V DC, two-wire 4–20 mA current loop Field of view / coverage Non-optical; pressure measured at the process port Measurement and monitoring component; replaces neither a pressure-relief valve nor an independent shutdown. Confirm compatibility with the actual cleaning chemical. Availability when checked Actively listed; DigiKey displayed stock when checked. Confirm quantity and delivery before ordering. Purchase / region ifm international; verified price is a US offer, check EU landed cost separately. - [ifm PT5400: manufacturer pressure and accuracy specifications ↗](https://www.ifm.com/gb/en/product/PT5400) - [DigiKey: exact PT5400 unit price and stock ↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/PT5400/12145157) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/PT5400/12145157) Liquid flow / water consumptionCompare: ifm SM6000 (feed line, maximum 16 bar) ifm ### SM6000 (feed line, maximum 16 bar) Monitoring conductive feed water at the ground station before the high-pressure pump 754.30 USDDigiKey US unit price; tax, shipping and any US tariffs added at checkout. Useful for consumption logging and insufficient feed-water detection when conductivity is sufficient. Monitor cleaning process #### Strengths - Measures flow, total volume and temperature in one industrial instrument. - No moving measuring parts; IO-Link provides process and diagnostic data. #### Weaknesses - Requires conductivity ≥20 µS/cm; very pure or demineralised cleaning water may be unsuitable. - The 16-bar limit and 544 g favour installation on the ground before the pump, not in the high-pressure line. IntegrationMedium IO-Link master or suitable pulse/current receiver at the ground controller; check water conductivity, 3×DN/1×DN inlet/outlet runs and a completely filled pipe. Operating range / quantity 0.1–25 l/min; maximum 16 bar. Suitable low-pressure feed line only, never a 100–250-bar cleaning line. Error / dispersion ±(0.8% of reading + 0.5% of range end value), manufacturer specification; does not prove nozzle flow if a downstream leak exists Interface IO-Link, switching/analogue/pulse outputs; G1/2 male DN15 process connection Specifications, availability & sources Rate / points Response 0.15 s at dAP=0/T19; a minimum 5 ms IO-Link process cycle does not mean the same hydraulic response Weight / housing 544 g; more appropriate for the ground station than the aircraft · IP67; wetted 316L stainless steel, PEEK and FKM Supply / power 18–30 V DC; data sheet states 95 mA at 24 V Field of view / coverage Non-optical; the DN15 measuring pipe must remain completely full Not for the high-pressure side, or for DI water without a conductivity check. Feed flow does not prove correct spray distribution or cleaning quality. Availability when checked Actively listed; DigiKey displayed stock when checked. Availability is not a delivery guarantee. Purchase / region ifm international; verified DigiKey price is in USD, local sourcing is possible. - [ifm SM6000: flow, conductivity, pressure limit and mass ↗](https://www.ifm.com/lt/en/product/SM6000) - [ifm SM6000: German manufacturer data sheet ↗](https://media.ifm.com/dam/35099e0b-39fb-4699-a406-9a31619d5a68/Original/SM6000-03_DE-DE.pdf) - [DigiKey: exact SM6000 unit price and stock ↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/SM6000/12144811) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.digikey.com/en/products/detail/ifm-efector-inc/SM6000/12144811) Force / hose tension (development setup)Compare: SparkFun SEN-21669 10 kg load cell + SEN-13879 HX711 SparkFun ### SEN-21669 10 kg load cell + SEN-13879 HX711 Low-cost hose-tension or reaction-force experiment, starting on a fixed test rig 24.90 USDSum of two separately ordered components: USD 19.95 + 4.95; tax and shipping extra. Not a preassembled manufacturer kit. Very inexpensive prototype for characterising hose tension; wet airborne use requires mechanical and electrical development. Monitor cleaning process #### Strengths - Both exact components have public individual prices. - A validated force trace can support hose management and warning thresholds. #### Weaknesses - An open hook and unverified attachment do not provide reliable flight or load retention. - One axis does not measure the full three-dimensional pull vector; side loads, vibration and temperature affect readings. IntegrationMedium Calibrate with known forces, limit side loading, design positive retention and an overload stop, and protect electronics from water. Start with data logging only. Operating range / quantity Rated load 10 kg, approximately 98 N at standard gravity; only in the intended mechanical loading direction Error / dispersion Manufacturer states 0.05% FS for the load cell; this does not automatically describe an installed, vibrating hose-tension measurement. Interface Four-wire strain-gauge bridge to HX711; clock/data to a microcontroller, no native CAN/PX4 Specifications, availability & sources Rate / points HX711 selectable 10 or 80 samples/s; measure filtering delay and oscillations in the complete assembly Weight / housing Combined cell, hook and board mass not stated on checked pages; mounting hardware is additional · No verified IP rating for the pair; exposed circuit board Supply / power HX711 2.7–5 V, <1.5 mA; cell excitation 3–12 V; use a common operating point within both ratings Field of view / coverage Non-optical; single-axis mechanical force measurement Development components, not certified load monitoring. The 10 kg model must suit measured operating pull and possible peaks. Availability when checked Both SparkFun pages displayed available stock when checked; order separately. Purchase / region US manufacturer with international electronics distribution; check import costs and delivery region. - [SparkFun SEN-21669: 10 kg hook load cell, price and specifications ↗](https://www.sparkfun.com/load-cell-10kg-straight-bar-with-hook.html) - [SparkFun SEN-13879: HX711 amplifier, price and specifications ↗](https://www.sparkfun.com/sparkfun-load-cell-amplifier-hx711.html) - [SparkFun: bridge wiring and calibration ↗](https://learn.sparkfun.com/tutorials/load-cell-amplifier-hx711-breakout-hookup-guide/introduction) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.sparkfun.com/load-cell-10kg-straight-bar-with-hook.html) Wind direction / wind speedCompare: Gill Instruments WindUltra 1416-PK-210-01 Gill Instruments ### WindUltra 1416-PK-210-01 Local wind and gust monitoring at an unobstructed ground station or on a validated aircraft boom Quote requiredNo reliable public price found for this exact configuration; request a quote including connection and mounting. Useful for weather assessment and logging; recommend initial installation at the ground station. Monitor cleaning process #### Strengths - Compact head with accuracy documented across its wind range. - No moving measuring components, diagnostic codes and an outdoor-rated design. #### Weaknesses - Rotor downwash and flow reflected from the wall distort onboard ambient-wind readings. - Ground wind cannot reliably represent gusts higher up or around a building edge. IntegrationHigh 1416-PK-210-01 includes head and M12 connector; add suitable pole/mounting/cable. Record timestamps and status; investigate motion, attitude and airflow errors for airborne installation. Operating range / quantity 0–75 m/s; direction 0–360°. Sensor range is not an approved wind limit for the cleaning aircraft. Error / dispersion 0–20 m/s: <2% RMSE or 0.1 m/s, whichever is greater; direction <1° RMSE. Above 20 m/s: <3% / <2°. RMSE is not a maximum-error guarantee. Interface RS485/RS422, Modbus RTU, NMEA and SDI-12; USB for configuration Specifications, availability & sources Rate / points Output 1, 2, 4 or 8 Hz; internal sampling 16 Hz Weight / housing 150 g sensor head; standard 1-inch pole mount adds 120 g, cable and boom additional · Manufacturer lists IP66, IP68 and IP69K; connector must be fully assembled Supply / power 5–30 V without heating; plan at least 9 V for full operating conditions. Unheated <0.8 W at high wind speeds, <1.5 W instantaneous peak; heated 2.4–7.6 W. Field of view / coverage Two-dimensional wind measurement in the sensor plane; not a 3D wind vector Does not authorise urban flight. Derive operating wind limits from the tested aircraft/hose system, not the sensor’s 75 m/s measuring range. Availability when checked Current Gill catalogue and manufacturer documentation; quote for price and lead time. Purchase / region UK manufacturer with international distribution; obtain a local quotation. - [Gill WindUltra: current manufacturer overview and accuracy ↗](https://gillinstruments.com/compare-2-axis-anemometers/windultra/) - [Gill WindUltra: Issue 2 manual, configurations, power and installation ↗](https://gillinstruments.com/windultra-1416-user-manual-issue-2/) - [Gill: full-range accuracy and RMSE explained ↗](https://gillinstruments.com/full-range-accuracy/) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://gillinstruments.com/compare-2-axis-anemometers/windultra/) 60 GHz radar development boardCompare: Acconeer XE125 with XM125 / A121 (evaluation board) Acconeer ### XE125 with XM125 / A121 (evaluation board) Research setup for radar distance/presence detection as an additional sensing principle 121.09 USDDigiKey US unit price; tax, shipping and any US tariffs added at checkout. Inexpensive radar experiment for a fixed rig; not yet wet-environment airborne distance protection. Surface / object recognition #### Strengths - XM125 with A121 and a microcontroller is already installed. - Manufacturer development tools and optional antenna/lens accessories. #### Weaknesses - Not a directly integrated PX4/Cube rangefinder with guaranteed complete-system accuracy. - Glass, multipath, wall angle and a wet enclosure require testing against the actual target. IntegrationHigh Develop profiles and target selection; flag dropouts/multiple targets, then build a timestamped companion interface. Include outdoor housing and radome in validation. Operating range / quantity Manufacturer states up to 20 m, depending on target, profile and installation; qualify minimum distance and a valid 50 cm working window for the application Error / dispersion No general facade-distance accuracy guaranteed for the complete XE125 system. Sensor technology alone does not establish ±5 cm flight regulation. Interface USB-UART and header I²C/UART; firmware/software bridge required Specifications, availability & sources Rate / points Profile and firmware dependent; no fixed complete-system rangefinder rate Weight / housing Complete XE125 mass not stated in the checked overview · Open evaluation PCB without a verified IP rating Supply / power 5 V USB-C or external header supply according to manual; profile and board peripherals determine consumption Field of view / coverage 60 GHz pulsed radar; separately sold lens/horn accessories can alter the field of view Development technology; does not replace a validated independent near-field sensor or pose estimator. Availability when checked Actively listed by Acconeer and DigiKey; DigiKey displayed stock when checked. Purchase / region Swedish manufacturer and international electronics distribution; check 60 GHz use and regional conformity of the final integration. - [Acconeer: XE125 / XM125 product overview ↗](https://acconeer.com/products/) - [DigiKey: exact XE125 variant and price ↗](https://www.digikey.com/en/products/detail/acconeer-ab/XE125/17883268) - [Acconeer XE125 hardware manual hosted by CODICO ↗](https://downloads.codico.com/misc/JKO/Acconeer/XE_XM125/XE125%20User%20Guide.pdf) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.digikey.com/en/products/detail/acconeer-ab/XE125/17883268) Radar altimeterCompare: Nanoradar NRA24 (current 200 m version) Nanoradar ### NRA24 (current 200 m version) Additional ground-clearance measurement and height reference, not a precision 50 cm wall controller Quote requiredCurrent 200 m configuration: quote required. Prices for older 50 m versions were not treated as the same product. Useful additional downward radar; lacks precision and near-range headroom for 45–55 cm facade control. Assessment at 50 cm sensor distance: Additional height measurement only: current ±10 cm and 0.5 m minimum range do not meet the 45–55 cm facade target. Height above ground #### Strengths - Enclosed radar altimeter with low power demand. - ArduPilot documents NRA24 support from firmware 4.5; verify the actual hardware/protocol variant. #### Weaknesses - ±10 cm sensor error already exceeds the overall ±5 cm facade target. - A 50 cm minimum range is at the bottom edge of the target band; vegetation, tilt and multipath affect the measured target. IntegrationMedium Check official ArduPilot CAN/UART documentation against the exact configuration; multiple NRA24 units on one CAN bus require unique IDs. Verify PX4 support separately. Operating range / quantity 0.5–200 m according to the current manufacturer page Error / dispersion Current manufacturer specification: ±0.1 m. Do not transfer older 50 m documents claiming 2 cm to the 200 m variant. Interface CAN / UART; confirm the firmware/protocol configuration being ordered Specifications, availability & sources Rate / points 42 Hz, current manufacturer page Weight / housing 95 g, current manufacturer page; older published mass values differ · IP66 per manufacturer; does not establish tolerance of direct pressure washing Supply / power 5–16 V DC; 1.5 W Field of view / coverage 28° azimuth × 18° elevation; beam and installation tilt determine the illuminated ground area Ground range does not provide complete local 3D pose or reliable window detection. Availability when checked Current manufacturer page lists the 200 m version; request price, firmware and delivery. Purchase / region Chinese manufacturer and international UAV distribution; confirm region, radio conformity and exact variant. - [Nanoradar: current NRA24 200 m specifications ↗](https://www.nanoradar.com/Products_UAV-Altimeter-Radar-NRA24.html) - [ArduPilot: official NRA24 integration and CAN limits ↗](https://docs.ardupilot.org/copter/docs/common-rangefinder-nra24.html) Checked: 23.10.2026 · Confirm variant, included items and delivery country with the supplier. [Product & price source↗](https://www.nanoradar.com/Products_UAV-Altimeter-Radar-NRA24.html) No sensors match this selection. Reset the search or filters. ### Direct comparison TURN A REQUIREMENT INTO A MEASUREMENT CONCEPT ## 50 cm target distance. ±5 cm for the complete system. 05 / 10 The nozzle should remain between 45 and 55 cm from the surface. The sensor contributes only part of the error: attitude, boom geometry, delays and response to wind, water-jet reaction or hose tension also matter. Sensor & Measurementpreferably ±1–2 cmGeometry, delay & Controlremaining margin Development target, with validation outstanding. For a conservative error estimate, relevant maximum error contributions add up; averaging does not remove unknown errors. - Resolution ≠ Accuracy. A sensor can report millimetres yet be centimetres off. - 1σ ≠ Guaranteed error bound. A dispersion figure does not bound every measurement error. - Nozzle ≠ Sensor. The positions and orientations of both components must be calibrated. - IP rating ≠ Measurement reliability. Droplets and mist can still interfere with a protected sensor. GEOMETRY CALCULATOR ### What must the sensor measure? Simplified planar facade. The sensor sits behind the nozzle; the offset is specified along the surface normal. Sensor behind nozzle 0 cmBeam angle to surface normal 0° Required beam distance50.0 cmMeasurement band for 45–55 cm nozzle distance: 45.0–55.0 cmBeam distance = (Nozzle distance + Offset) / cos(Angle) Geometric illustration. The angle must be known; boom deflection, sensor errors and motion are excluded. 0 cm / 0°: 50 cm. 60 cm / 0°: 110 cm. 01 ### Learn to measure at low cost URM14 for glass experiments and an optical cross-check sensor. TFmini Plus is inexpensive but its ±5 cm leaves no margin for the complete system. 02 ### Compare measurement principles URM14 + LW20/HA: USD 488 in listed sensor prices before extra costs. Different measurement principles are an advantage; fault tolerance can only be assessed after testing. 03 ### Add 3D sensing with a purpose Compare MID-360S or Airy as mapping / odometry candidates, with Hesai and Ouster as industrial references. A high LiDAR price does not establish precise nozzle distance. When sensors disagree: Do not blindly average readings. An optical sensor might see through a window. Evaluate measurement age, signal quality, viewing direction and the surface model; do not enable water while the state is uncertain. Plan the flight response for the facade, hose and surroundings. Recommendation inferred from specifications. Ultrasound can detect transparent surfaces; smooth surfaces remain angle-dependent. [OMRON: ultrasound for transparent objects and temperature effects ↗](https://www.ia.omron.com/support/guide/55/further_information.html) · [Pepperl+Fuchs: ultrasonic technology guide ↗](https://files.pepperl-fuchs.com/webcat/navi/productInfo/doct/tdoct3557a_eng.pdf?v=20231211180838) Local navigation · without satellites or RTK corrections ## Can the drone know its position without RTK? Yes: a local coordinate system can be sufficient for a facade. The drone still needs continuous measurements of its motion or pose relative to fixed references. The ground station's coordinate alone does not provide that measurement. Operation without an RTK base is feasible; permanently exact position without any reference is not a defensible promise. ### Why a remote-controlled car works without GPS For an ordinary RC car, a person watches the vehicle and corrects steering and speed. The car does not need to know its own coordinates in metres. A person can also pilot a drone without GPS. Automatic hovering and repeatable cleaning paths, however, need an estimate of motion and position. Wheels and the ground support a car; a drone must additionally control altitude and every flight movement. Attitude means tilt and orientation, altitude is a vertical coordinate, and position consists of three spatial coordinates. An IMU helps stabilize attitude, but simply integrating its readings does not maintain accurate position indefinitely. A barometer or a single wall-distance measurement does not provide a full 3D position either. [Position reference basics](https://www.bitcraze.io/documentation/system/positioning/) ### Building coordinates instead of latitude and longitude A useful working frame can start at a surveyed building corner: u runs along the facade, v points upwards and d points perpendicularly away from the facade plane. A target can be “u = 3 m, v = 8 m, nozzle distance = 0.50 m”. These coordinates do not need satellites. Fixed markers, anchors or a recognized map link the estimate to the building; a start-relative estimate alone can drift over time. A rangefinder solves only part of the problem: many locations along and above a facade have the same wall distance. An angled sensor also measures its beam path, not automatically the perpendicular separation. Sensor, IMU and nozzle positions and vehicle attitude must agree geometrically. Transform these working axes into the coordinate frame expected by the flight controller. | Method | Output / product examples | Strengths | Facade limitations | Stereo/VIO + IMU | Local pose and velocity from image motion and IMU measurements. Cameras: [ZED X Mini], [OAK-D Pro](https://shop.luxonis.com/products/oak-d-pro); compute platform: [VOXL 2]. | No installed radio anchors required; motion estimation runs onboard. | A camera alone is not a complete VIO system. Glass, repeated windows, weak texture, backlight, droplets and vibration can disrupt tracking; drift remains without fixed relocalization. [VIO practice](https://docs.modalai.com/flying-with-vio/) | LiDAR-inertial odometry / SLAM | 3D LiDAR, IMU and software register the environment and estimate motion. Examples: Livox MID-360 or Ouster OS0 with a compatible LIO pipeline. | Uses spatial geometry even when visible image texture is weak. | A large flat wall provides little information about motion parallel to it; glass and spray remain difficult. A point cloud alone is not vehicle position; drivers, timing and calibration are required. [FAST-LIO](https://github.com/hku-mars/FAST_LIO) | Optical flow + ground range | Image flow and range constrain horizontal velocity. Examples: Holybro H-Flow or ARK Flow. | Compact assistance for low hover above suitable ground. | No fixed building reference or drift-free absolute position. Ground texture, lighting and usable ground range limit altitude; viewing a wall sideways is not a direct replacement for a downward-facing integration. [PX4 optical flow](https://docs.px4.io/main/en/sensor/optical_flow) | Fixed AprilTag / ArUco markers | A calibrated camera estimates pose relative to markers of known size and surveyed building position. | Affordable fixed references; can correct VIO drift and support relocalization. | Requires visible, surveyed, weather-resistant targets. Resolution, angle and range determine performance. Multiple markers help consistency checks; there is no universal centimetre guarantee. [AprilTag](https://april.eecs.umich.edu/software/apriltag) · [ArUco](https://docs.opencv.org/4.x/d5/dae/tutorial_aruco_detection.html) | UWB anchors + tag | Radio ranges to local anchors determine position; for example [Nooploop LinkTrack P-B]. | No GNSS or RTK corrections; fixed local reference. | One ordinary ranging anchor provides only a distance. Robust 3D positioning requires multiple well-distributed, calibrated anchors, including elevation diversity. Ground-only anchors often provide weak height geometry at greater distances. Metal and multipath can bias ranges. LinkTrack's typical 30 cm 3D figure cannot alone meet a ±5 cm nozzle target. [Manufacturer data](https://www.nooploop.com/linktrack/) | Ultrasonic beacons | Acoustic time of flight between fixed beacons and a mobile receiver; e.g. [Marvelmind Industrial Super-MP-3D]. | Local position without satellites; a different measurement principle from camera/LiDAR. | Requires clear acoustic paths and suitable geometry. Wind, rotors and spray need testing. Beacons determine position between devices, not the surface of a window. | External cameras / total station | External cameras track a rigid marker body, e.g. [OptiTrack PrimeX 13]. A [Leica TS16](https://leica-geosystems.com/en-gb/products/total-stations/robotic-total-stations/leica-ts16) can track a prism for reference experiments. | Independent local measurement; little airborne electronics. | Occlusion, latency, outdoor light and network connectivity limit use. One prism does not give full orientation. A total station is not a flight controller; validate update rate and dynamic tracking separately. | Standalone GNSS | Global coordinates without a local correction base. | Still useful for coarse outdoor orientation. | It remains satellite-dependent. Building occlusion and reflections prevent a defensible ±5 cm facade-control guarantee; it does not meet a satellite-free requirement. ### Which architecture would I build? #### 1 · Affordable first experiment: assistance A pilot flies along the facade. A suitable ultrasonic sensor, such as URM14 or an industrial microsonic, monitors local separation; a second measurement principle checks consistency on suitable surfaces. H-Flow/ARK Flow can assist close to suitable ground. This stage does not know dependable metre-based building coordinates or execute an autonomous cleaning path. #### 2 · Local paths: VIO + fixed markers For a bounded test area: Cube/Pixhawk with IMU, compatible VIO computer and tracking camera, several surveyed AprilTags, and an independent wall rangefinder. Markers reference u/v and orientation; VIO bridges short gaps. Downward range helps only within its actual ground range. This would be my preferred next development step where marker installation is possible. #### 3 · Larger work area: multiple measurement methods Combine LiDAR-inertial navigation and VIO; use recognizable facade features or markers to reference the fixed map. Optionally add UWB with anchors at different elevations as an additional position aid. Control precise nozzle separation independently. External cameras can measure a test rig independently but need a new deployment design for mobile facade work. These tiers are development proposals. No listed component establishes trouble-free cleaning flight or a permanent ±5 cm total outdoor error. Tag and camera prices do not include complete infrastructure systems. ### What belongs at the ground station, and what stays onboard? The ground station can manage the origin, building map, cleaning paths and telemetry. External cameras may also compute measurements there. Fast attitude and flight-control loops stay in the flight controller. A timestamped local position estimate must reach it reliably; a map point or PC-generated target does not replace a measured current position. Lost connectivity or invalid localization needs a previously tested response. Blindly commanding “hold position” cannot work when the position estimate itself has failed. PX4: supply external pose preferably using MAVLink ODOMETRY with correct frames, timestamps and uncertainties. Documentation specifies 30–50 Hz for external vision; configure EKF2_EV_CTRL, sensor offsets and delay for the actual firmware. ArduPilot: ODOMETRY is the preferred ExternalNav message; VISION_POSITION_ESTIMATE is also supported. The documented minimum 4 Hz is an interface requirement, not a recommended rate for precise facade control. An EKF origin may be required without GPS; entering it does not create a position measurement. [PX4 integration](https://docs.px4.io/main/en/ros/external_position_estimation) · [ArduPilot ExternalNav](https://ardupilot.org/dev/docs/mavlink-nongps-position-estimation.html) Measure before autonomous flight: verify known u/v/d locations, return along the same path and measure residual error. Log dropouts, drift per time/distance, end-to-end latency, relocalization and resets; then add rotor wash and water. Only the measured performance of the complete setup establishes readiness for local cleaning paths. M17 · ARDUPILOT / SIYI N7 · SENSORS AND COMPUTE ## Which combination fits your budget? 07 / 10 Dynamic hardware suggestions for your existing M17. Around 10 kg additional payload is a provisional user estimate; actual flying mass, N7 port use and power supply still need confirmation. Planning, not a binding full cost estimate: tax-unconfirmed shop prices are not confirmed net quotes. The default 19% uplift is an adjustable budgeting assumption. Accessories and reference materials are estimates; verify shipping, import, software licensing, development, testing and operational costs separately. Aircraft, flight controller, pump and hose system are excluded. Price review: 23 October 2026. Exchange rate: ECB 22 October 2026, EUR 1 = USD 1.1205; not a bank or checkout rate. [ECB source](https://www.ecb.europa.eu/stats/policy_and_exchange_rates/euro_reference_exchange_rates/html/eurofxref-graph-usd.en.html) · [Planner data JSON] ### Our recommendation for city-centre exterior façades For your team, I would develop supervised automatic coverage within a controlled work area. Your existing SIYI N7 running ArduPilot handles flight control; a companion computer estimates local position and plans overlapping cleaning passes. The ground station defines the façade, exclusions and operating boundaries. It cannot replace continuous position measurements on the aircraft. This is a proposed development architecture, rather than an already demonstrated autonomous cleaning system. #### Solve two measurement tasks separately Distance from the wall: For a 50 cm nozzle standoff, I would first compare the [microsonic pico+100/I] with the less expensive [DFRobot URM14]. Ultrasound can detect glass, although angled glass, rotor airflow and water can still cause invalid echoes. The The budget packages use LightWare LW20/HA; the DT35 below is an alternative to cost separately. [SICK DT35-B15251] is an optical cross-check candidate on suitable opaque surfaces. Readings that pass through glass must not be averaged blindly with ultrasound. Sensor location, nozzle offset and aircraft attitude belong in the geometry; 50 cm nozzle standoff does not establish rotor clearance. Multiple ultrasonic sensors need coordinated measurement cycles to avoid interference. Position along the wall: A distance reading cannot identify the current floor or the pass already cleaned. This needs a position and orientation estimate with known uncertainty. The [ArduPilot External Navigation documentation](https://ardupilot.org/dev/docs/mavlink-nongps-position-estimation.html) describes external pose/velocity input. Coordinate frames, timestamps, missing data and position resets must match the selected firmware. GNSS/RTK can supplement navigation in open sky; obstruction and reflected signals near façades can degrade it. I would not use it as the sole wall reference. #### Practical development stages | Stage | Products | What to establish first | Measurement rig | URM14 or pico+100/I; DT35 for optical comparison | Error, dropouts and latency on dry/wet glass, render and metal; then add rotor airflow and spray. | Controlled flight prototype | [OAK-D Pro W, OV9782] + Seeed reComputer Super J3011; VIO software, surveyed AprilTags and independent wall sensor | Local VIO position, reference reacquisition and small automatic cleaning areas under pilot supervision. | Larger façade areas | [Livox MID-360S] or [RoboSense Airy 96]; companion computer, IMU/SLAM processing and references | Environment mapping, motion estimation and recorded coverage; cameras/markers help on geometrically repetitive walls. As an alternative VIO platform, the linked [VOXL 2] price covers the board only; cameras and integration are additional. AprilTags require accurate printed dimensions, surveyed positions and appropriate visibility during flight. VIO drifts between references. [ModalAI documents VIO with AprilTag relocalisation](https://docs.modalai.com/voxl-vision-hub-apriltag-relocalization/); that does not establish centimetre accuracy on your building. The demonstrated workflow requires separate integration and validation in your SIYI N7/ArduPilot stack. LiDAR has similar scene limitations: a long flat wall provides little information about movement along its plane. [Livox also specifies near-range limitations](https://www.livoxtech.com/mid-360s/specs). A 3D sensor therefore does not replace independent wall-distance measurement. #### Measure the cleaning process and operating limits Process monitoring should include pressure, for example the [ifm PT5400], and a suitably specified flow meter. The [ifm SM6000] is an example for the low-pressure side before the pump: maximum 16 bar and minimum conductivity of 20 µS/cm. It is often unsuitable for deionised cleaning water. Hose tension requires separate force measurement or validated hose-system monitoring. Pressure and flow alone cannot establish correct nozzle standoff or an unobstructed spray path. Test window edges, cables, protrusions, water films, gusts and spray/hose forces. Cameras do not guarantee person detection. If position or distance quality becomes insufficient, a tested response must manage water release and flight state. A low-cost 2D scanner observes a single plane; indoor motion capture needs suitable camera visibility and a prepared capture volume. Both can support development, but neither is a universal city-centre solution. The budgets above cover selected sensors and computers. Accessories and reference materials are allowances; the tax uplift is an assumption. Check actual mounts, power conversion, enclosures and adapters against these allowances. Software engineering, licences, validation, shipping and import are not fully costed. A larger budget does not purchase approved autonomy. Clarify the operating category early for spraying in a city: the open category prohibits dropping material, and EASA distinguishes automatic routes with pilot intervention from autonomous operation. See [EASA: specific category](https://www.easa.europa.eu/en/domains/drones-air-mobility/operating-drone/specific-category-civil-drones) and [EASA: automatic/autonomous](https://www.easa.europa.eu/en/the-agency/faqs/drones-uas). The initial development target is a controlled area, a remote pilot present and demonstrated intervention capability. SIYI N7 specifics: The official board description maps TELEM1 to SERIAL1 and GPS2/TELEM4 to SERIAL4; SERIAL2 is unused. Confirm occupied ports and installed firmware first. A companion needs its own suitable power supply; UART signals use 3.3 V. The budget includes the computer and keeps the existing flight controller. [Official N7 board description](https://github.com/ArduPilot/ardupilot/blob/master/libraries/AP_HAL_ChibiOS/hwdef/SIYI_N7/README.md) FROM THE SIX ATTACHED DOCUMENTS ## Assess research, preserve knowledge. 08 / 10 The three autonomy updates are preserved as a dated research archive. All 15 linked arXiv titles were checked. Titles and abstracts match; detailed figures were checked against full text for selected papers only. Particularly relevant · GLIO2 ### Real cleaning exposes a perception limit. Ten facade missions: eight completed autonomously; two needed manual takeover after water spray was detected as an obstacle. Reported UAV-to-wall distance was 1–2.5 m. This does not demonstrate 50 cm nozzle distance with ±5 cm maximum deviation. [GLIO2 · Full text ↗](https://arxiv.org/html/2610.12411v1) 16.10.2026Monitor sensor quality and validate real safety margins5 papers New papers examine LiDAR / radar switching, edge SLAM and safe quadrotor navigation. Smoke tests on ground robots and idealised perception do not directly establish performance on wet glass. For DroneShine, quality monitoring, timestamps, reliable pump shutdown and reproducible disturbance tests remain priorities. Research / PreprintTest an offline quality monitor #### [When LiDAR Fades, Radar Takes Over: LiDAR Reliability-Aware Adaptive Localization Under Airborne Obscurants ↗](https://arxiv.org/abs/2610.13788) A quality monitor switches to radar odometry when LiDAR degrades. The full text confirms four ground-robot tunnel runs with artificial smoke. Limit: No independent ground truth: reported RMSE measures agreement between estimators. Smoke trials do not demonstrate water-mist performance; planar output only. Research / PreprintMonitor code release #### [LIN-SLAM: Lightweight LiDAR-Inertial Neural SLAM for Real-Time Onboard Navigation ↗](https://arxiv.org/abs/2610.21199) Hybrid geometric frontend and neural map backend; the abstract confirms real-time operation on Unitree Go2 with Jetson Orin NX over 1.47 km. Limit: Ground robot and repetitive corridors, rather than wet-facade cleaning. The abstract announces code after acceptance. Research / PreprintAdopt the validation approach #### [Constrained Differentiable Policy Learning for Safe Quadrotor Navigation ↗](https://arxiv.org/abs/2610.16119) Explicit safety constraints during training; the abstract confirms transfer to Gazebo / PX4 and a real drone. Limit: No guarantee for DroneShine safety distances. Check simulation margins against real perception errors, delays and disturbance forces. Research / PreprintMonitor #### [Safety-Critical Quadrotor Navigation Using Fourth-Order Control Barrier Functions with Full Nonlinear Quadrotor Dynamics ↗](https://arxiv.org/abs/2610.14522) A safety filter includes full rigid-body dynamics and acts at torque level; simulation and real experiments are confirmed in the abstract. Limit: Deep controller integration rather than a simple offboard retrofit. No demonstrated operation with water jets, hoses or degraded onboard perception. Research / PreprintAssess behaviour rules #### [ClearNav: Perception-Aligned Quadrotor Flight Using Differential-Physics ↗](https://arxiv.org/abs/2610.17419) Navigation aligns camera and motion directions and slows down near obstacles; simulated and real tests are confirmed. Limit: No demonstrated suitability for glass / spray or complete PX4 / ArduPilot integration. Invalid depth must be treated as unknown in the actual system. 09.10.2026Real facade cleaning shows limits of spray perception5 papers GLIO2 demonstrates ten cleaning missions at 1–2.5 m UAV-to-wall distance; eight finish autonomously, two need manual takeover after spray produces false obstacles. Radar odometry and simulation are supplementary research components and do not replace precise surface-relative ranging. Research / PreprintArchitecture benchmark #### [GLIO2: A GPU-Parallelized Tightly-Coupled LiDAR-Inertial-GNSS System for Robust and Real-Time Global Localization and Mapping ↗](https://arxiv.org/abs/2610.12411) LiDAR, IMU and GNSS are tightly coupled. The full text confirms ten facade-cleaning missions: eight autonomous, two takeovers due to water spray interpreted as obstacles; UAV-to-wall distance 1–2.5 m. Limit: No numerical demonstration of 50 cm ±5 cm distance control; research stage with announced code. The 110 m figure is building height, not flight distance. Research / PreprintRedundant diagnostic channel #### [Embedded Bare-Metal Radar-Inertial Odometry ↗](https://arxiv.org/abs/2610.07278) Radar-inertial estimation on a microcontroller; the abstract confirms real flight with unmodified PX4 and 0.51–0.82 m translational APE. Limit:Odometry in this error range does not demonstrate centimetre-level distance control. No confirmed tests with water mist, wet glass or a cleaning hose. Research / PreprintMonitor #### [NMPP: Nonlinear Model Predictive Planning for Agile UAV Flight in Cluttered Environments ↗](https://arxiv.org/abs/2610.08695) Dynamically executable flight planning with perceived obstacles as hard geometric constraints; the abstract confirms real flights up to 5.5 m/s. Limit:No surface-coverage control or demonstrated cleaning with hose and spray disturbances. Research / PreprintMonitor simulation #### [UWB Meets Crazyflow: Simulating Degraded Feedback at Scale for Aerial Robotics ↗](https://arxiv.org/abs/2610.08225) UWB / IMU simulation and a complete EKF in a differentiable JAX simulator; a modular approach for multiple failure scenarios. Limit:Simulation results do not replace PX4 / ArduPilot SITL / HITL or actual DroneShine disturbance-force measurements. Research / PreprintLong-term research #### [Predicting Cable Dynamics with Physical Attention Bias ↗](https://arxiv.org/abs/2610.11975) Geometrically informed attention improves prediction of unseen cables in simulation. Limit:Short simulated cables are not a validated model of a long, filled and pressurised cleaning hose. 02.10.2026Perceive glass, compare sensors, model load changes5 papers GlassGuard adds glass surfaces to maps; L1-MPPI explores adaptive control; ForVis demonstrates the effect of sensor selection. Results come from different research setups and do not establish complete cleaning with a wet facade, hose tension and a water jet. Research / PreprintOffline candidate #### [GlassGuard: Verified Glass Plane Mapping for Robot Navigation ↗](https://arxiv.org/abs/2610.02110) Visual glass segmentation and LiDAR geometry add missing glass planes to a navigation map. The abstract confirms nine building scenes and 82% coverage with identical pinhole inputs. Limit:Research mapping component; no demonstrated distance control on wet glass or in spray. A public repository exists, but the licence is unclear. Research / PreprintMonitor SITL / HITL #### [L1-MPPI: L1 Adaptive Model Predictive Path Integral for Agile UAV Control ↗](https://arxiv.org/abs/2609.38467) Adaptive control improves trajectory tracking with unknown additional load. The abstract confirms real flights with 35% added mass. Limit:Rigid mass is not a filled hose. No confirmed suitability for 50 cm target distance with ±5 cm maximum deviation. Research / PreprintMonitor #### [LiDARFlow: Real-Time Panel-Based MAV Guidance in Unknown Environments ↗](https://arxiv.org/abs/2610.01573) Lightweight reactive obstacle guidance from local LiDAR measurements; the abstract confirms two real indoor scenarios. Limit:Obstacle guidance does not replace localisation or precise facade control. Detailed figures and autopilot porting need separate checks. Research / PreprintBenchmark reference #### [ForVis: An In-Field Dataset and Benchmark for VIO Using Under-Canopy UAV Flights in Forests ↗](https://arxiv.org/abs/2609.35482) Twelve UAV flights and 504 evaluations of seven open VIO methods show substantial differences between camera / IMU sensors. Limit:A forest benchmark does not prove performance near wet, reflective facades. Useful as a template for collecting comparison data. Research / PreprintCurrently low priority #### [Degeneracy-Orthogonal Geometric Constraints for LiDAR SLAM ↗](https://arxiv.org/abs/2609.36753) DeCOD uses recognisable cross-sectional structures for drift correction in tunnels and pipes. Limit:A flat glass facade generally does not meet those geometric assumptions. Three additional foundations from the attachments - [Autonomous Close-Proximity Photovoltaic Panel Coating Using a Quadcopter ↗](https://arxiv.org/abs/2509.10979) Onboard VIO, PV-corner detection and model-based compensation for ground effect and decreasing liquid mass. The full text confirms outdoor z-RMSE of 8.9 cm without and 3.0 cm with compensation. RMSE is not maximum error or a ±3 cm guarantee. A 1.56 kg prototype with 150 ml of liquid and a D455; not a high-pressure / hose-cleaning system. - [FAST-LIO2: Fast Direct LiDAR-inertial Odometry ↗](https://arxiv.org/abs/2107.06829) Direct LiDAR-inertial odometry with raw-point registration; the abstract supports UAV / ARM operation and an open-source implementation. Foundational component; specific sensor, glass and spray suitability requires separate tests. - [Survey on LiDAR Perception in Adverse Weather Conditions ↗](https://arxiv.org/abs/2304.06312) Survey of LiDAR degradation in fog, rain and snow, including processing and sensor fusion. General overview, not a product comparison or validation of a particular radar / LiDAR safety channel on a cleaning drone. Earlier claims corrected “±5–10 mm wall measurement”, “±1–3 cm actual flight distance” and “OS0 is the best device” are not measured DroneShine properties. A second optical channel is not automatically independent on glass or in mist. A paper's RMSE is not a guaranteed maximum deviation. FROM DATASHEET TO OWN VALIDATION ## A manageable development path 09 / 10 - 01 ### Record the requirement and geometry Is distance defined from the nozzle tip or sensor? Record offset, beam angle, working speed, pressure, hose state, temperature and material. Define acceptance criteria for maximum deviation, outage duration and shutdown latency before testing. - 02 ### Fixed rig: 40 / 50 / 60 cm Measure render, dry glass, wet glass and PV surfaces against an independent reference. Start perpendicular, then vary defined angles. Compare dry and wet sensor windows and ultrasonic membranes. - 03 ### Add airflow and spray separately Test airflow first, then mist, then their combination reproducibly. Log signal quality, timestamps, dropouts and false targets; a real pressure jet needs a suitable protected test rig. - 04 ### Test fusion and failure responses Inject stale, frozen and conflicting readings. Do not average valid and invalid measurements. Water enable, valve response, computer restart and recovery must work traceably. - 05 ### Simulation, limited flight, then cleaning Simulate control and faults first. Then use a suitable protected flight setup with independent distance measurement. Qualify hose tension, jet reaction and wind separately and together before expanding the operating envelope. Maximum errorlargest nozzle deviation Outliers / Dropoutsfrequency and duration Measurement age / Latencyuntil water shutdown RMSE as a supplementdoes not replace maximum error TRACEABLE AND EXTENSIBLE ## Sources & Knowledge maintenance 10 / 10 ### How claims are classified - Manufacturer data: Range, interface, IP rating and error specifications under the stated test conditions. - Shop price: Linked variant, currency and tax status as of the entry’s review date; not an automatic live price. - Research: Dated papers with experimental limitations, not product approval. - Development target / Inference: Proposed architecture, selection and test plan; own validation remains outstanding. ### Create complete new entries Manufacturer, model and revision; product and datasheet URLs; price, currency, tax status and date; range, accuracy, temperature, rate, weight, power and interface; strengths, weaknesses, integration effort, availability, plus own dry and wet test results. Checked data is also available as JSON. Prices and availability are snapshots; confirm region and exact included items with the supplier. New findings are maintained in the shared catalogue. [Download product JSON ↓][Download complete Markdown overview ↓] Official documentation and access to shared chats - [ArduPilot: Rangefinder ↗](https://ardupilot.org/copter/docs/common-rangefinder-landingpage.html) - [PX4: ROS 2 User Guide ↗](https://docs.px4.io/main/en/ros2/user_guide) - [QGroundControl: User Guide ↗](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/) - [Mission Planner: documentation ↗](https://docs.ardupilot.org/planner/index.html) - [Shared chat: original hardware discussion ↗](https://chatgpt.com/share/6ada6b5b-4024-83eb-8c64-83ed833e6a89) - [Shared chat: additional discussion ↗](https://chatgpt.com/share/6ada6b88-5ea0-83eb-89d2-e5dbcdc66812) The two shared pages could not be read when fetched. The fully pasted chat and all six attachments were used; shared-page content is not presented as additionally verified. DroneShine · Control & SensingAs of 23 October 2026 · Engineering knowledge with sources[Back to top ↑] ## Budget planner: M17 / ArduPilot / SIYI N7 [Interactive budget selection](https://drone.office-dashboard.com/roadmap/wissen/en/#budgetberater) BOM totals are provisional hardware planning figures: confirmed net prices and tax-unconfirmed USD shop prices are labelled separately. 19% is an optional German budget uplift, not a binding landed-price calculation. Unconfirmed does not mean net. Exchange, shipping, import and checkout may differ. Aircraft, autopilot, RC/ground station, pumps, water circuit, hose management, software development, licenses, authorization and operational controls are excluded. Accessory allowances are estimates rather than vendor quotes. ### Sensor bench only Material tests against glass, render and wet surfaces. No automatic flight path. - 1 × [DFRobot URM14 SEN0358](https://www.dfrobot.com/product-2173.html): 99 USD / unit; unconfirmed Accessory allowance: 80 EUR; marker materials: 0 EUR. Fixed bench; no flight-control integration, 3D pose or optical cross-check. ### Piloted distance assistance Develop and log two different wall-ranging methods. The pilot commands vertical and lateral movement. - 1 × [DFRobot URM14 SEN0358](https://www.dfrobot.com/product-2173.html): 99 USD / unit; unconfirmed - 1 × [LightWare LW20/HA](https://nwblue.com/products/lw20-ha-100-m): 389 USD / unit; unconfirmed Accessory allowance: 120 EUR; marker materials: 0 EUR. Do not classify this as an autonomous cleaning drone. Conflicting or invalid readings must cause a pause; do not average through-glass readings. ### Local paths with markers and VIO Develop repeatable local paths on a prepared facade: camera/IMU for motion, surveyed markers for reference, ultrasound plus laser for wall ranging. - 1 × [microsonic pico+100/I](https://sensorpartners.com/en/product/microsonic-pico100-i/): 234.1 EUR / unit; net - 1 × [LightWare LW20/HA](https://nwblue.com/products/lw20-ha-100-m): 389 USD / unit; unconfirmed - 1 × [Luxonis OAK-D Pro W USB, OV9782 RGB](https://shop.luxonis.com/products/oak-d-pro-w): 579 USD / unit; unconfirmed - 1 × [Seeed reComputer Super J3011, SKU 114110312](https://www.seeedstudio.com/reComputer-Super-J3011-p-6444.html): 769 USD / unit; unconfirmed - 1 × [ifm PT5400 · 0–400 bar](https://www.digikey.com/en/products/detail/ifm-efector-inc/PT5400/12145157): 176.1 USD / unit; unconfirmed Accessory allowance: 400 EUR; marker materials: 80 EUR. Requires visible, sufficiently distributed markers and tested pose/failsafe software. No guaranteed ±5 cm. Software integration, commercial VIO licensing and surveying labor are not priced. ### Preferred 3D development setup 3D LiDAR/IMU for surroundings and LIO, camera/IMU as a second modality, two time-coordinated ultrasonic heads and an optical wall channel. Markers provide a surveyed reference when needed. - 2 × [microsonic pico+100/I](https://sensorpartners.com/en/product/microsonic-pico100-i/): 234.1 EUR / unit; net - 1 × [LightWare LW20/HA](https://nwblue.com/products/lw20-ha-100-m): 389 USD / unit; unconfirmed - 1 × [Luxonis OAK-D Pro W USB, OV9782 RGB](https://shop.luxonis.com/products/oak-d-pro-w): 579 USD / unit; unconfirmed - 1 × [Seeed reComputer Super J3011, SKU 114110312](https://www.seeedstudio.com/reComputer-Super-J3011-p-6444.html): 769 USD / unit; unconfirmed - 1 × [Livox MID-360S](https://tpi-livox.eu/products/mid360s.html): 599 EUR / unit; net - 1 × [Livox Connector Cable 1.5 m, MID-360S](https://tpi-livox.eu/products/mid360s.html): 79 EUR / unit; net - 1 × [ifm PT5400 · 0–400 bar](https://www.digikey.com/en/products/detail/ifm-efector-inc/PT5400/12145157): 176.1 USD / unit; unconfirmed Accessory allowance: 650 EUR; marker materials: 120 EUR. Two ultrasound heads can share environmental failure modes and require synchronization. Flat, smooth or repetitive facades can make LIO/VIO degenerate despite a higher price. Without markers or other local references, absolute repeatability is not assured. ### Conditional wider-FOV 3D comparison Airy replaces MID-360S and its cable in this comparison. The larger vertical field may improve coverage of overhangs and upper/lower surroundings, depending on mounting. - 2 × [microsonic pico+100/I](https://sensorpartners.com/en/product/microsonic-pico100-i/): 234.1 EUR / unit; net - 1 × [LightWare LW20/HA](https://nwblue.com/products/lw20-ha-100-m): 389 USD / unit; unconfirmed - 1 × [Luxonis OAK-D Pro W USB, OV9782 RGB](https://shop.luxonis.com/products/oak-d-pro-w): 579 USD / unit; unconfirmed - 1 × [Seeed reComputer Super J3011, SKU 114110312](https://www.seeedstudio.com/reComputer-Super-J3011-p-6444.html): 769 USD / unit; unconfirmed - 1 × [RoboSense Airy Device Only 96-Beam](https://store.robosense.ai/products/airy?variant=44401434165301): 999 USD / unit; unconfirmed - 1 × [ifm PT5400 · 0–400 bar](https://www.digikey.com/en/products/detail/ifm-efector-inc/PT5400/12145157): 176.1 USD / unit; unconfirmed Accessory allowance: 850 EUR; marker materials: 120 EUR. Not an automatic purchase recommendation: availability and final tax treatment are unclear. Wider field does not make glass or spray reliable and does not deliver validated autonomous cleaning.