Journey into ROVR - LightCone: The 3D Data Engine Driving Generative AI
Last updated
Last updated
The ROVR LightCone device is the world’s first 3D data collection platform that combines advanced automotive-grade LiDAR technology, the GEODNET global RTK network, and powerful computing capabilities to deliver high-precision, high-value data for industries such as robotics, autonomous driving, and intelligent transportation. Whether in urban streets or remote areas, the LightCone ensures stable, accurate data collection, accelerating the development of next-generation generative AI.
High-Definition Cameras: While the LiDAR provides high-precision ranging and geometric data, the pre-calibrated HD camera enhances the LiDAR point cloud by adding valuable semantic information.
Global High-Precision Positioning: Powered by the GEODNET global RTK network, the LightCone device delivers centimeter-level positioning accuracy on a global scale.
Efficient Data Processing: Equipped with a high-performance edge processor, the device processes large-scale data efficiently, minimizing computational latency and ensuring real-time data handling.
Comprehensive Environmental Adaptability: Operating in temperatures ranging from -45°C to 85°C, with an IP67 waterproof rating, the LightCone device is designed to perform reliably in extreme weather conditions, including high and low temperatures, rain, and snow. As an active sensor, the LiDAR ensures effective operation even in low-light and nighttime environments.
High Stability and Reliability: Constructed with automotive-grade sensors and components, the LightCone device provides exceptional durability and long-term reliability, ensuring stable performance in complex environments.
Easy Installation and Operation: Featuring a detachable design with a separate base and main unit, the LightCone device allows for easy installation and removal. Its compact and flexible design ensures compatibility with a variety of vehicles, enabling rapid deployment without the need for complex setup.
LiDAR Type
Auto-grade solid-state LiDAR
Number of Beams
126-beam
Ranging Accuracy
±5cm
Eye Safety
IEC60825-1 Class 1
Measurement Range
200m (180m @10% NIST)
Blind Spot
≤0.5m
Field of View
120° * 25°
Angular Resolution
0.2° * 0.2°
Camera
Auto-grade HD camera
Positioning Accuracy
±3cm @ global, powered by GEODNET
IMU Bias Instability (Allan)
Gyro: XY < 2.6°/hour, Z < 2.1°/hour Accelerometer: < 45µg
IMU Bias (XYZ)
0.13°/√h, 0.050m/√h
Processor
High-performance edge processor
NPU
6TOPs @ int8
GPU
ARM Mali-G610MC4
Working Temperature
-45°C ~ 85°C
Storage Temperature
-65°C ~ 105°C
Data Storage
1 TB
Weight
2.3kg + 1.6kg
Dimensions
251.6mm * 204.3mm * 96.2mm
Protection Rating
IP67
Items
ROVR-LightCone
Google Map Vehicle
Google Map Backpack Device
Applicable Scenarios
Vehicle-mounted Outdoor
Vehicle-mounted Outdoor
Backpack Indoor/Outdoor
Integration
Integrated
Split-type
Integrated
Dimensions
251.6mm * 204.3mm * 96.2mm
Approximately the same size as the vehicle
Similar size to LC
LiDAR Beam Count (Data Density)
126 beams
Typically 40 beams
Less than 40 beams
Scalability
Equipped with high-performance edge computing chip
Only a data collection device
Only a data collection device
Ingress Protection Rating
IP67
N/A
N/A
Price
$2,500
$200,000
$20,000
LiDAR Type
Automotive-grade LiDAR
360° LiDAR
360° LiDAR
Design Life
10-15 years
<5 years
<5 years
Sensor Calibration Difficulty
Pre-calibrated
Calibration required each time
Pre-calibrated
Ease of Use
No specialized training required
Requires professional personnel
Requires professional personnel
LC data has broad applications across various high-tech industries, including autonomous driving, robotics, simulation platforms, and more. Based on different processing methods, LC data is categorized into raw data and processed data, with specific applications as follows:
Autonomous Driving Systems: LC's raw data and derived data are valuable for perception and model training in autonomous driving systems. These data help autonomous driving technology companies improve their perception capabilities and algorithm accuracy. The processed high-definition map data offers precise navigation and decision-making support, ensuring safe driving in complex road conditions and dynamic environments.
Robotics: LC's raw data can be used in the perception systems of various robots (such as industrial robots, service robots, drones, etc.), helping robots understand and adapt to their surroundings. This improves their autonomous navigation, obstacle avoidance, and task execution abilities. Processed data can provide precise environmental modeling and path planning support, optimizing robots' operational performance.
Simulation Platforms: LC data is essential for providing accurate 3D scenes and environmental data for simulation platforms used in autonomous driving and robotics. This aids in algorithm validation, testing, and optimization. Whether it's road scenes, building structures, or natural environments, the high-precision data from the LC device can be used to create highly realistic simulation environments, accelerating technology iteration and validation.
Data Annotation Services: The high-precision 3D data collected by LC offers abundant raw data for data annotation service providers. This enables them to supply the labeled data needed for training autonomous driving and robotics algorithms, helping clients improve the accuracy and stability of their models in real-world applications.
If you are interested in purchasing the ROVR LightCone device or would like to learn more about the technical details, feel free to contact us. Our team will provide you with comprehensive technical support and services.
Thank you!
ROVR Team
🔸 ROVR Network ()
💬 X: ROVR ()
🗯 Discord ()
