RoboSense and BitRobot have formed a strategic partnership focused on 3D perception and real-world data collection. A key outcome of this collaboration is the RoboCap LiDAR-100 dataset, released by BitRobot. Built on RoboSense's 3D perception solution, the dataset features 100 hours of real-world operational scenarios—including kitchen, auto repair, and auto detailing settings—making it the world's first 3D manipulation dataset with high-resolution, accurate depth measurement. The dataset can simultaneously record LiDAR, RGB imagery, and inertial measurement data, delivering QVGA-resolution depth information that provides precise spatial data for hand tracking, head pose estimation, and environment reconstruction. This empowers robots to understand the world and learn manipulation skills through high-quality multimodal data. BitRobot is an open R&D network for robotics focused on embodied intelligence. It has previously co-released a humanoid robot dataset with Unitree Robotics and Hugging Face, and its team has also contributed to robot evaluation research alongside Google DeepMind and OpenAI. This collaboration extends RoboSense's 3D perception solution into the physical AI training data production stage, advancing its related products from robot body perception toward the physical AI data entry point.