DEP-RL: Embodied Exploration for Reinforcement Learning in Overactuated and Musculoskeletal Systems
This addresses a key bottleneck in applying reinforcement learning to complex, muscle-actuated systems, which could advance robotics and biomechanics.
The paper tackled the problem of ineffective exploration in reinforcement learning for overactuated musculoskeletal systems, achieving fast learning of reaching and locomotion tasks with improved sample efficiency and robustness compared to current approaches.
Muscle-actuated organisms are capable of learning an unparalleled diversity of dexterous movements despite their vast amount of muscles. Reinforcement learning (RL) on large musculoskeletal models, however, has not been able to show similar performance. We conjecture that ineffective exploration in large overactuated action spaces is a key problem. This is supported by the finding that common exploration noise strategies are inadequate in synthetic examples of overactuated systems. We identify differential extrinsic plasticity (DEP), a method from the domain of self-organization, as being able to induce state-space covering exploration within seconds of interaction. By integrating DEP into RL, we achieve fast learning of reaching and locomotion in musculoskeletal systems, outperforming current approaches in all considered tasks in sample efficiency and robustness.