Industry Research

Appearance

RoboSense LiDAR Technical Report ​

RSHELIOS and RSBP Models for Airside Autonomous Vehicle Operations ​

Last updated: 2026-03-22

1. Company Overview ​

RoboSense Technology Co., Ltd. ​

Founded: 2014, Shenzhen, China

Founders: Qiu Chunxin (CEO, PhD from Harbin Institute of Technology on outdoor robotics perception), Zhu Xiaorui (Chief Scientist, PhD supervisor to Qiu), and Liu Letian (CTO, PhD peer).

Stock Listing: Listed on the Hong Kong Stock Exchange on 5 January 2024 under ticker 2498.HK -- the first IPO on HKEX in 2024 and the first "laser radar stock" in the Hong Kong market. IPO raised HK$985.12 million (~US$126 million) by offering 22.9 million shares at HK$43 each. Initial market capitalisation exceeded HKD 19 billion, making RoboSense the world's largest LiDAR company by market value at the time.

Key Investors: Over 30 institutional investors including Cainiao (Alibaba's logistics arm, 10.46% pre-IPO stake), BYD, and Xiaomi.

IPO Proceeds Allocation: 45% R&D, 20% manufacturing/testing/verification, 20% sales/marketing, 15% partnerships and working capital.

2024 Financial Performance:

  • Total revenue: ~RMB 1.65 billion (YoY +47.2%, three consecutive years of high growth)
  • Total LiDAR units sold: ~544,000 (YoY +109.6%)
  • Overall gross margin: ~17.2% (Q4 2024 gross margin: 22.1%)
  • Adjusted net loss: RMB 396 million (YoY improvement of 8.9%)
  • Global automotive LiDAR market share: 33.5% (No. 1 globally per Gasgoo Research Institute)

Milestone: In February 2025, RoboSense rolled off its 1-millionth LiDAR unit (an E1R), becoming the first company globally to deliver 1 million high-resolution LiDAR units.

Core Technology Differentiator: First LiDAR company to build its own chip technology, including self-developed RISC-V data processing SoCs, digital large-area SPAD-SoCs, 2D addressable VCSELs, and 2D MEMS scanning devices.

2. RS-Helios 32-Channel Series (RSHELIOS) ​

The Helios series is the successor to the RS-LiDAR-32, offering 29% smaller form factor and 60% lower cost compared to the original RS-LiDAR-32.

2.1 Model Nomenclature ​

The Helios 32 model numbers encode the vertical FOV:

  • RS-Helios-5515 (a.k.a. Helios-32 F70): -55 deg to +15 deg = 70 deg total vertical FOV
  • RS-Helios-1615 (a.k.a. Helios-32 F31): -16 deg to +15 deg = 31 deg total vertical FOV
  • Helios-32 F26: -16 deg to +10 deg = 26 deg total vertical FOV

2.2 Common Specifications (All Helios 32 Variants) ​

ParameterSpecification
Channels32
Wavelength905 nm
Laser SafetyClass 1 (IEC 60825-1), eye-safe
Horizontal FOV360 deg
Horizontal Angular Resolution0.1 deg / 0.2 deg / 0.4 deg (selectable)
Range150 m maximum; 90 m @ 10% NIST reflectivity
Near-Field Blind Spot<= 0.2 m (5515 variant: <= 0.1 m)
Range Accuracy+/-3 cm (0.1-1 m); +/-2 cm (1-100 m); +/-3 cm (100-150 m)
Point Rate (Single Return)576,000 pts/s
Point Rate (Dual Return)1,152,000 pts/s
Rotation Speed300 / 600 / 1200 rpm
Frame Rate5 / 10 / 20 Hz
Data Interface100Base-T1 Ethernet (automotive-grade)
Output ProtocolUDP packets (MSOP + DIFOP)
Input Voltage9-32 V DC
Power Consumption12 W
Dimensionsdia.100 mm x H 100 mm
Weight~1.0 kg (without cabling)
Operating Temperature-40 deg C to +60 deg C
Storage Temperature-40 deg C to +85 deg C
IP RatingIP67 and IP6K9K
Return ModesSingle return, Dual return

2.3 Variant-Specific Differences ​

ParameterHelios-5515 (F70)Helios-1615 (F31)Helios F26
Vertical FOV70 deg (-55 to +15)31 deg (-16 to +15)26 deg (-16 to +10)
Vertical Angular ResolutionUp to 1.33 deg1.0 deg (uniform)Up to 0.5 deg
Beam DistributionNon-uniform (dense centre, sparse edges)UniformNon-uniform (dense centre)
Primary Use CaseNear-field + blind-spot detectionSurveying, mapping, uniform coverageLong-range perception, highest resolution
Near-Field Blind Spot<= 0.1 m<= 0.2 m<= 0.2 m

2.4 Beam Pattern Details ​

Helios-5515 (F70): Arranges dense laser beams in the middle of the 70 deg vertical FOV and sparse beams at both ends. The 55 deg downward tilt greatly reduces the near-field blind zone, making it ideal for low-speed autonomous vehicles that need to detect ground-level obstacles close to the vehicle.

Helios-1615 (F31): Uniform 1 deg vertical spacing across 31 deg. Provides consistent angular density, preferred for surveying and mapping applications where uniform point cloud density matters.

Helios F26: Highest vertical angular resolution at 0.5 deg within 26 deg FOV. Non-uniform distribution with denser beams concentrated in the central FOV region for maximum object discrimination at range.

2.5 Operating Modes ​

  • High-performance mode: Full point rate at maximum rotation speed
  • Low power consumption mode: Reduced rotation speed and power draw
  • Web configuration and monitoring: Browser-based configuration interface
  • Multi-radar interference shielding: Built-in protection against cross-talk from adjacent LiDAR units

3. RS-Bpearl (RSBP) 32-Channel Specifications ​

The RS-Bpearl is a hemispherical-FOV LiDAR designed specifically for near-field and blind-spot detection around autonomous vehicles and robots.

3.1 Specifications ​

ParameterSpecification
Channels32
Wavelength905 nm
Laser SafetyClass 1, eye-safe
Horizontal FOV360 deg
Vertical FOV90 deg (hemispherical)
Horizontal Angular Resolution0.2 deg (10 Hz) / 0.4 deg (20 Hz)
Vertical DistributionNon-uniform (32 channels across 90 deg)
Range30 m @ 10% reflectivity
Near-Field Blind Spot< 10 cm (~0 blind spot)
Range Accuracy+/-3 cm (typical)
Point Rate (Single Return)576,000 pts/s
Point Rate (Dual Return)1,152,000 pts/s
Frame Rate10 / 20 Hz (600 / 1200 rpm)
Data Interface100 Mbps Ethernet
Input Voltage9-32 V DC
Power Consumption~13 W (typical)
Dimensionsdia.111 mm x H 100 mm
Weight~0.92 kg
Operating Temperature-40 deg C to +60 deg C
Storage Temperature-40 deg C to +85 deg C
IP RatingIP67
Return ModesSingle return, Dual return

3.2 Key Differences: RSBP vs RSHELIOS ​

AspectRS-Bpearl (RSBP)RS-Helios (RSHELIOS)
RoleNear-field / blind-spot detectionPrimary perception sensor
Vertical FOV90 deg (hemispherical)26-70 deg (forward-looking)
Range30 m150 m
Accuracy+/-3 cm+/-2 cm (1-100 m range)
Use CaseGround obstacles, curbs, close pedestriansLong-range object detection, path planning
MountingTypically roof-centre or bumper cornersRoof-top or mast-mounted
Weight0.92 kg1.0 kg
Power~13 W12 W

The RSBP and RSHELIOS are complementary sensors. In autonomous vehicle deployments, the RSBP handles the immediate surround while the RSHELIOS provides the long-range forward/360 deg perception layer.

4. RS-LiDAR-M Series (MEMS Solid-State, ASIL-B) ​

4.1 Technology ​

The M series uses RoboSense's proprietary 2D MEMS smart scanner chips, advancing from 1D mechanical scanning to 2D chip-based scanning. No motors, ball bearings, or wear surfaces exist in the optical path. This solid-state architecture enables:

  • Lower cost at scale
  • Higher reliability (no mechanical wear)
  • Automotive-grade durability

4.2 M1 Plus Key Specifications ​

ParameterSpecification
Range200 m max; 180 m @ 10% NIST
FOV120 deg (H) x 25 deg (V)
Angular Resolution0.2 deg x 0.1 deg (standard)
Smart GAZE ROI Resolution0.1 deg (V) in dynamically selected ROI
Point Rate (Dual Return)Up to 1,575,000 pts/s
Input Voltage9-16 V
Power Consumption15 W

4.3 Functional Safety: ASIL-B ​

RoboSense adheres strictly to ISO 26262 safety standards for the M series:

  • Random hardware failure rate: < 10^-7 /h (fully achieving ASIL-B requirements)
  • Functional safety level: ASIL-B (also SIL-2 for industrial)
  • Integrated fail-safe concepts from aerospace and rail transportation
  • Safety mechanism covers thousands of failure modes across:
    • Laser emitter and receiver monitoring
    • MEMS control monitoring
    • Point cloud processing and transmitting monitoring

4.4 Automotive-Grade Certifications ​

  • MEMS mirror module: AEC-Q100 certification (reliability test report by SGS)
  • Eye safety: IEC 60825-1 Class 1 (certified by SGS and Goebel)
  • Automotive-grade test standards applied:
    • ISO 16750 (road vehicle environmental testing)
    • GB/T18655-2010 / CISPR 25:2008 (EMC)
    • ISO 11452 (EMC component test)
    • ISO 7637 (electrical disturbances)
    • ISO 10605 (ESD)
    • IEC 60068 (environmental testing)

4.5 Reliability Test Data (M1 Platform) ​

TestDuration/Result
High-temperature durability> 36,000 hours
High-humidity testing> 24,000 hours
Cyclic temperature shock> 21,000 hours
Cumulative test time (all samples)> 300,000 hours
Longest continuous prototype operation> 700 days
Total road test mileage> 200,000 km

4.6 M Series Product Range ​

  • M1: Original solid-state MEMS LiDAR (SOP announced CES 2021)
  • M1 Plus: Enhanced range (200 m) and Smart GAZE function
  • M2: Mid-range, 200 m @ 10%, 0.1 deg x 0.1 deg ROI resolution
  • M3: Next-generation variant
  • MX: Extended variant

5. rslidar_sdk ROS/ROS2 Driver ​

5.1 Overview ​

The rslidar_sdk is the official ROS/ROS2 driver for all RoboSense LiDAR products. It wraps the core rs_driver library and provides standard ROS integration.

Repository: https://github.com/RoboSense-LiDAR/rslidar_sdkLatest Release: v1.5.18 (15 July 2025), 560 commits, 12 contributors

5.2 Supported Models ​

The driver supports 18 LiDAR types via the lidar_type YAML parameter:

RS16, RS32, RSBP, RSAIRY, RSHELIOS, RSHELIOS_16P, RS128, RS80, RS48,
RSP128, RSP80, RSP48, RSM1, RSM1_JUMBO, RSM2, RSM3, RSE1, RSMX

5.3 ROS/ROS2 Compatibility ​

PlatformROS Version
Ubuntu 16.04ROS Kinetic
Ubuntu 18.04ROS Melodic / ROS2 Eloquent
Ubuntu 20.04ROS Noetic / ROS2 Galactic
Ubuntu 22.04ROS2 Humble

5.4 PointCloud2 Fields: XYZIRT ​

The driver supports two point types, configured via POINT_TYPE in CMakeLists.txt:

XYZI (basic):

cpp
struct PointXYZI {
  float x;      // metres
  float y;      // metres
  float z;      // metres
  uint8_t intensity;  // 0-255
};

XYZIRT (full, recommended):

cpp
struct PointXYZIRT {
  float x;           // metres
  float y;           // metres
  float z;           // metres
  uint8_t intensity; // 0-255
  uint16_t ring;     // channel/beam ID (0-31 for 32ch sensors)
  double timestamp;  // per-point timestamp (seconds)
};

When published as ROS sensor_msgs/PointCloud2, the fields map to:

FieldPointCloud2 TypeOffset
xFLOAT320
yFLOAT324
zFLOAT328
intensityFLOAT32 (cast from uint8)12
ringUINT1616
timestampFLOAT6418

The ring field is critical for algorithms that need per-beam processing (ground segmentation, beam-specific calibration). The timestamp field provides per-point timing for motion compensation during ego-motion.

5.5 Key Configuration Parameters (config.yaml) ​

yaml
common:
  msg_source: 1                    # 1=online LiDAR, 2=ROS packet, 3=PCAP
  send_packet_ros: false
  send_point_cloud_ros: true

lidar:
  - driver:
      lidar_type: RSHELIOS         # LiDAR model identifier
      msop_port: 6699              # MSOP data port
      difop_port: 7788             # DIFOP status port
      min_distance: 0.2            # Minimum range filter (m)
      max_distance: 200.0          # Maximum range filter (m)
      use_lidar_clock: true        # Use LiDAR hardware clock vs host
      dense_points: false          # Exclude NaN points if true
      ts_first_point: true         # Timestamp = first point or last
      start_angle: 0               # Start angle of scan (deg)
      end_angle: 360               # End angle of scan (deg)
    ros:
      ros_frame_id: rslidar
      ros_send_point_cloud_topic: /rslidar_points

5.6 Dependencies ​

  • libyaml-cpp-dev (>= 0.5.2)
  • libpcap-dev (>= 1.7.4)
  • PCL (included with ROS desktop-full)

5.7 Build and Launch ​

ROS (catkin):

bash
cd ~/catkin_ws
catkin_make
source devel/setup.bash
roslaunch rslidar_sdk start.launch

ROS2 (colcon):

bash
cd ~/ros2_ws
colcon build
source install/setup.bash
ros2 launch rslidar_sdk start.py

5.8 Velodyne Compatibility ​

An open-source tool rs_to_velodyne (https://github.com/HViktorTsoi/rs_to_velodyne) converts RoboSense point clouds to Velodyne format, enabling use of existing Velodyne-based perception pipelines without modification.

6. Time Synchronisation: PTP/gPTP ​

6.1 Supported Synchronisation Methods ​

RoboSense LiDAR supports multiple time synchronisation approaches:

MethodProtocol LayerPrecisionNotes
GPS/GNSS (PPS + GPRMC)Hardware pulse + serialMicrosecond-levelIndustry standard, requires GPS antenna
PTP (IEEE 1588)L2 EthernetSub-microsecondLower cost than GPS, no antenna needed
gPTP (IEEE 802.1AS)L2 EthernetSub-microsecondRequires hardware timestamp support

6.2 PTP Implementation Details ​

  • RoboSense PTP uses L2 (Ethernet layer) for communication
  • Supports Peer-to-Peer (P2P) delay measurement mechanism
  • PTP precision: sub-microsecond level
  • gPTP shares the same Peer Delay Mechanism with PTP but has stricter hardware requirements (hardware timestamps mandatory)

6.3 Timestamp Architecture ​

  • Timestamps stored in MSOP packet headers (bytes 21-30, within 42-byte header)
  • Temporal resolution: microsecond-level (1 us theoretical minimum)
  • Per-block timestamps at ~111 us granularity (at 600 RPM)
  • use_lidar_clock parameter in rslidar_sdk selects between LiDAR hardware clock and host system clock

6.4 Synchronisation for Multi-Sensor Fusion ​

For airside AV deployments requiring camera-LiDAR-IMU fusion:

  • PTP/gPTP over automotive Ethernet is the recommended approach (no GPS antenna required indoors/under jetways)
  • PTP master clock should be the central compute unit or a dedicated grandmaster
  • All sensors on the same PTP domain achieve sub-microsecond alignment
  • The timestamp field in XYZIRT point clouds enables per-point motion compensation

7. Adverse Weather Performance ​

7.1 Built-In Weather Filtering ​

The Helios and Bpearl series include Rain, Fog, Snow, and Dust Denoising functions (available upon request / firmware configuration). These operate at the sensor firmware level to filter weather-related noise from the point cloud before transmission.

7.2 Dual-Return Mode for Weather ​

In dual-return mode, each laser pulse registers two range returns. This is critical in adverse weather:

  • First return: May hit a raindrop, snowflake, or fog particle
  • Second return: Penetrates to the actual surface behind the particle
  • Enables downstream algorithms to identify and discard weather artifacts while retaining true object detections

7.3 Quantitative Weather Degradation (General 905 nm LiDAR) ​

ConditionTypical Impact
Light rain (< 2.5 mm/h)Minimal degradation, denoising effective
Heavy rain (> 7.5 mm/h)Max detection range decreases ~30%, point density drops ~45%
Moderate fog (visibility 200-500 m)Reduced range, increased noise
Dense fog (visibility < 200 m)Significant range reduction, heavy noise
Moderate snowDenoising maintains clear point cloud (validated at -23 deg C)
DustAddressed by denoising function

7.4 Validated Cold-Weather Testing (M1) ​

RoboSense conducted the first cold-winter test of automotive-grade solid-state LiDAR in northeast China (Yakeshi, Inner Mongolia and Heihe, Heilongjiang):

  • Temperature range: -18 deg C to -23 deg C
  • Conditions: Low-visibility moderate snow, Level 4-5 northwest winds
  • Result: "Reliable perception performance with clear and stable point cloud output"
  • The RS-Helios series is rated to -40 deg C operating temperature, providing additional cold-weather margin

7.5 IP Protection Summary ​

ModelIP RatingSignificance
RS-Helios 32IP67, IP6K9KDust-tight, submersion-proof, high-pressure jet wash
RS-BpearlIP67Dust-tight, submersion to 1 m for 30 min
RS-LiDAR-M1/M1+IP6K9KAutomotive-grade pressure-jet water resistance

IP6K9K is particularly important for airside operations where vehicles are exposed to jet wash, de-icing fluid spray, and pressure cleaning.

8. Comparison: RoboSense Helios 32 vs Hesai XT32 ​

ParameterRS-Helios-1615 (F31)Hesai XT32
Channels3232
Wavelength905 nm905 nm
Range (max)150 m120 m
Range @ 10% NIST90 m80 m
Horizontal FOV360 deg360 deg
Vertical FOV31 deg31 deg
H Angular Resolution0.1/0.2/0.4 deg0.18 deg
V Angular Resolution1.0 deg1.0 deg
Point Rate (Single)576,000 pts/s640,000 pts/s
Range Accuracy+/-2 cm (1-100 m)+/-1 cm
Range Precision--0.5 cm (1-sigma)
Weight~1.0 kg~0.8 kg
Dimensionsdia.100 x 100 mmdia.76 x 103.2 mm
Power Consumption12 W~18 W (typical)
Operating Temp-40 to +60 deg C-20 to +40 deg C
IP RatingIP67, IP6K9KIP6K7
Data Interface100Base-T1100 Mbps Ethernet
Design LifespanNot published> 30,000 hours (typical)
Eye SafetyClass 1Class 1
Dual ReturnYes (1,152,000 pts/s)Yes (1,280,000 pts/s)
Weather DenoisingBuilt-in firmware functionNot specified
Near-Field Blind<= 0.2 m0.05 m
Approx. Price~US$1,800-2,700~US$3,000-4,000

8.1 Key Advantages: RoboSense Helios ​

  • Wider operating temperature range (-40 to +60 vs -20 to +40) -- critical for airside operations across seasons
  • Superior IP rating (IP6K9K vs IP6K7) -- jet-wash and pressure-cleaning resilient
  • Greater range (150 m vs 120 m max)
  • Lower power consumption (12 W vs ~18 W)
  • Built-in weather denoising in firmware
  • Multiple FOV variants (26/31/70 deg) from same platform
  • Selectable horizontal resolution (0.1/0.2/0.4 deg)
  • Lower cost (~40-50% less expensive)

8.2 Key Advantages: Hesai XT32 ​

  • Better range accuracy (+/-1 cm vs +/-2 cm)
  • Higher point rate in single return (640k vs 576k pts/s)
  • Lighter weight (0.8 kg vs 1.0 kg)
  • Smaller diameter (76 mm vs 100 mm)
  • Shorter near-field blind spot (0.05 m vs 0.2 m)
  • Published design lifespan (> 30,000 hours)

9. Reliability Data ​

9.1 Mechanical LiDAR (Helios/Bpearl) ​

RoboSense Helios and Bpearl are designed with reference to automotive-grade standards, with reliability tests covering:

  • Mechanical shock
  • Random vibration
  • Low-temperature operation (-40 deg C)
  • Water protection (IP67/IP6K9K)
  • EMC (electromagnetic compatibility)

Specific MTBF figures for the Helios and Bpearl mechanical LiDAR units are not publicly disclosed. However, the automotive-grade design and testing regime implies reliability comparable to automotive-tier components.

9.2 Solid-State LiDAR (M Series) ​

The M platform has the most comprehensive published reliability data:

  • Random hardware failure rate: < 10^-7 /h (ASIL-B compliant)
    • This equates to a theoretical MTBF > 10,000,000 hours
  • AEC-Q100 certification for MEMS mirror module
  • 300,000+ hours cumulative test time
  • 700+ days longest continuous operation of a single prototype
  • 200,000+ km total road test mileage
  • Testing per ISO 16750, ISO 11452, ISO 7637, ISO 10605, IEC 60068

9.3 E1R Solid-State (Latest Generation) ​

  • Passed over 60 rigorous reliability tests
  • Operating temperature: -40 to +85 deg C
  • Vibration shock tolerance: up to 50 G
  • Fully solid-state (no moving parts) -- inherently longer lifespan

10. Optimal Airside Configuration ​

10.1 Airside Environment Characteristics ​

Airport airside environments present specific challenges:

  • Wide open aprons with long sight-line requirements (> 100 m)
  • Mixed traffic: aircraft, baggage tugs, fuel trucks, ground crew on foot
  • Jet blast and jet wash exposure
  • FOD (Foreign Object Debris) detection requirements at ground level
  • All-weather operation (rain, snow, fog, extreme temperatures)
  • GPS-denied areas under terminal buildings, jetways, and hangars
  • High-vibration environments (vehicle traversing expansion joints, rough tarmac)
  • Pressure washing and de-icing fluid exposure

10.2 Recommended Sensor Configuration ​

Primary Perception Layer: 2x RS-Helios-1615 (F31)

  • Mount: Roof-top, front and rear facing
  • Configuration: 31 deg vertical FOV with uniform 1 deg resolution
  • Role: Long-range (150 m) 360 deg perception, path planning, obstacle detection
  • Rationale: Uniform beam distribution preferred for consistent point cloud density at all ranges; 150 m range covers full apron crossing distances; -40 deg C rating handles all climatic conditions

Near-Field / Blind-Spot Layer: 4x RS-Bpearl

  • Mount: Four corners of the vehicle (front-left, front-right, rear-left, rear-right), tilted slightly outward
  • Configuration: 90 deg hemispherical FOV, < 10 cm blind spot
  • Role: Ground-level obstacle detection, curb detection, FOD identification, close-proximity pedestrian safety
  • Rationale: Hemispherical coverage eliminates blind spots around the vehicle; 30 m range sufficient for safety-critical close-range envelope

Optional Forward Perception Enhancement: 1x RS-LiDAR-M1 Plus

  • Mount: Front-centre, forward-facing
  • Configuration: 120 deg x 25 deg FOV, 200 m range, ASIL-B
  • Role: Long-range forward detection on taxiways, enhanced resolution in ROI via Smart GAZE
  • Rationale: ASIL-B functional safety for safety-critical forward detection; solid-state reliability in high-vibration environment

10.3 Configuration Parameters for Airside ​

Horizontal Resolution: Set to 0.2 deg for Helios units (optimal balance of point density and data bandwidth at 10 Hz frame rate).

Frame Rate: 10 Hz recommended. Provides adequate temporal resolution for low-speed airside vehicles (typically < 25 km/h) while keeping data bandwidth manageable.

Dual-Return Mode: Enable for all sensors. Critical for adverse weather operations -- dual return allows perception algorithms to see through rain, snow, and dust.

Weather Denoising: Enable firmware-level rain/fog/snow/dust denoising on all Helios and Bpearl units.

Time Synchronisation: Use PTP over the automotive Ethernet backbone. PTP is preferred over GPS for airside because:

  • GPS signals may be occluded under terminal buildings, jetways, and inside hangars
  • PTP provides sub-microsecond synchronisation without antenna placement constraints
  • All sensors share the same PTP domain via the vehicle Ethernet switch

rslidar_sdk Configuration:

  • Set use_lidar_clock: true to use hardware-synchronised timestamps
  • Set dense_points: true to exclude NaN points, reducing downstream processing load
  • Use XYZIRT point type for full per-point ring and timestamp data
  • Set min_distance to 0.2 m for Helios, 0.1 m for Bpearl

10.4 Data Architecture ​

Total sensor data budget (recommended configuration):

SensorCountPoints/s (Dual Return)Total
RS-Helios-161521,152,0002,304,000
RS-Bpearl41,152,0004,608,000
RS-LiDAR-M1 Plus11,575,0001,575,000
Total7~8.5 M pts/s

Each XYZIRT point = 26 bytes, so total raw bandwidth = ~221 MB/s. This is well within the capacity of a GbE backbone with dedicated VLAN per sensor.

10.5 Mounting and Environmental Considerations ​

  • IP6K9K rating on the Helios units means they survive airside pressure-washing operations without removal or bagging
  • The 905 nm wavelength is eye-safe at Class 1, important for operations around ground crew and passengers
  • 9-32 V input range accommodates both 12 V and 24 V vehicle electrical systems common in ground support equipment
  • At ~1 kg per Helios and ~0.92 kg per Bpearl, total LiDAR weight for the 7-sensor configuration is approximately 5.7 kg -- negligible for a ground vehicle

11. Additional RoboSense Product Lines (Reference) ​

E1R (Solid-State, Robotics) ​

  • 120 deg x 90 deg ultra-wide FOV
  • 30 m @ 10% range, 75 m max
  • 144-beam, 260,000 pts/s (single) / 520,000 pts/s (dual)
  • 0.625 deg angular resolution
  • Digital SPAD-SoC + 2D VCSEL chips
  • -40 to +85 deg C operating
  • 50 G vibration shock
  • 69.5 x 95 x 43 mm, very compact

EM4 (Thousand-Beam Digital LiDAR) ​

  • Integrates SPAD-SoC with 1080-Core LEP (LiDAR Echo Processing)
  • Proprietary "Large Echo Processing Model" (Huiyan AI Model)
  • AI-trained noise reduction for rain, fog, and dust
  • Next-generation digital architecture

Sources ​

Public research notes collected from public sources.

Copyright © 2026 kvynlim