Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper investigates the impact of passive body dynamics on bipedal robot locomotion using model-based deep reinforcement learning. They trained two simulated bipedal robots, one with passive elements (springs) and one without, finding that the passive robot's training was initially slower but ultimately yielded more robust and energy-efficient locomotion due to attractor-driven learning. The key result is that incorporating passive elements allows robots to exploit stable limit cycles for high-performance locomotion.
Passive elements in robot design aren't just about hardware simplicity; they unlock a fundamentally more robust and energy-efficient approach to learning locomotion.
Embodiment is a significant keyword in recent machine learning fields. This study focused on the passive nature of the body of a biped robot to generate walking and running locomotion using model-based deep reinforcement learning. We constructed two models in a simulator, one with passive elements (e.g., springs) and the other, which is similar to general humanoids, without passive elements. The training of the model with passive elements was highly affected by the attractor of the system. This lead that although the trajectories quickly converged to limit cycles, it took a long time to obtain large rewards. However, thanks to the attractor-driven learning, the acquired locomotion was robust and energy-efficient. The results revealed that robots with passive elements could efficiently acquire high-performance locomotion by utilizing stable limit cycles generated through dynamic interaction between the body and ground. This study demonstrates the importance of implementing passive properties in the body for future embodied AI.