LanderPi Demonstrates Multimodal Embodied Robotic Autonomy

The LanderPi project introduces a multimodal composite robot that fuses large language models, 3D vision, LiDAR and motion control to interpret natural language and execute physical tasks. Using a 3D structured-light camera, YOLOv11 for edge detection, inverse kinematics for a 6-DOF arm and onboard planning, LanderPi locates, grasps and tracks objects in cluttered environments.