Gasgoo Munich- RoboSense's 3D sensing hardware was used to collect the real-world data included in RoboCap LiDAR-100, a new robotics dataset released by BitRobot during CoRL 2026. The dataset contains 100 hours of workplace recordings and is now available on Hugging Face.
Image source: RoboSense
Key Takeaways
- RoboSense 3D sensing hardware was used to capture 100 hours of data from restaurant kitchens, automotive repair facilities and car-detailing operations.
- BitRobot unveiled RoboCap LiDAR-100 during CoRL 2026 as an outcome of its collaboration with RoboSense.
- The dataset is now available on Hugging Face.
- RoboSense and BitRobot plan to explore further integration between 3D sensing hardware and open robotics development networks.
RoboSense's role in data collection
RoboSense develops robotics technologies for physical AI, including internally developed chips and digital perception products. The company positions these technologies as core components and infrastructure for robots operating in real-world environments.
According to RoboSense, its digital perception hardware is designed to provide data for robotic environmental perception, spatial understanding and skill learning. In the RoboCap LiDAR-100 project, its 3D sensing hardware was used to collect data from complex workplace environments.
BitRobot's open robotics network
BitRobot operates an open robotics research and data network for embodied AI. Its activities cover real-world data collection, robot teleoperation and physical-robot evaluation, with the aim of testing robotics technologies outside laboratory settings.
BitRobot said its internally developed data-collection hardware has entered volume delivery and that it is pursuing open-data collaborations with partners including Hugging Face and Unitree Robotics. It is also one of the initiators of RoboColosseum, an open program for evaluating robot models in real-world environments.
What to watch next
RoboSense and BitRobot plan to continue exploring how 3D sensing hardware can be integrated into open robotics development networks. The companies said the collaboration will seek to make real-world data more accessible to developers and support further advances in robotic perception and learning.
