ProDapt: Proprioceptive Adaptation using Long-term Memory Diffusion

Schoellig 1, 3 Abstract -- Diffusion models have revolutionized imitation learning, allowing robots to replicate complex behaviours. However, diffusion often relies on cameras and other exteroceptive sensors to observe the environment and lacks long-term memory. In space, military, and underwater applications, robots must be highly robust to failures in exteroceptive sensors, operating using only proprioceptive information. In this paper, we propose ProDapt, a method of incorporating long-term memory of previous contacts between the robot and the environment in the diffusion process, allowing it to complete tasks using only proprioceptive data. This is achieved by identifying "keypoints", essential past observations maintained as inputs to the policy. We test our approach using a UR10e robotic arm in both simulation and real experiments and demonstrate the necessity of this long-term memory for task completion. I NTRODUCTION Imitation learning aims to learn a control policy directly from observing others perform a task. However, imitation learning faces many challenges, including high-dimensional observation and action spaces, multi-modality, and strong temporal correlation. These challenges have limited the performance of imitation learning on complex and precise tasks. Diffusion has been successfully applied to imitation learning [1], [2].