RoboSense, BitRobot Forge Strategic Partnership for Physical AI Data Collection

RoboSense announced the partnership on November 11 at CoRL 2026, stating that the two companies will combine 3D perception hardware and real-world data collection to support embodied AI development. The collaboration aims to make physical-world data easier to capture and use for robot learning.

BitRobot's RoboCap LiDAR-100 dataset is the first public outcome. It uses RoboSense's 3D perception solution to record 100 hours of real work in kitchens, auto repair shops and car detailing studios, capturing multimodal data from everyday tasks rather than controlled laboratory settings. The companies say such data can help robots understand complex environments and learn manipulation skills.

For the physical AI ecosystem, the partnership points to a data-entry layer: perception sensors and structured datasets can feed training pipelines for embodied models. By focusing on messy, real operational scenes, RoboCap LiDAR-100 is designed to support researchers testing whether models generalize beyond clean benchmarks and into commercial service and industrial settings.