Robust Humanoid Contact Planning with Learned Zero- and One-Step Capturability Prediction
This work addresses the problem of robust contact planning for humanoid robots in dynamic environments, representing an incremental improvement over existing reactive methods.
The paper tackles the problem of humanoid robots failing to recover from external disturbances due to limited contact options, by proposing a search-based footstep planner that maximizes the probability of reaching a goal without falling. The results show that this approach generates more robust footstep sequences than a conventional planner in four challenging scenarios.
Humanoid robots maintain balance and navigate by controlling the contact wrenches applied to the environment. While it is possible to plan dynamically-feasible motion that applies appropriate wrenches using existing methods, a humanoid may also be affected by external disturbances. Existing systems typically rely on controllers to reactively recover from disturbances. However, such controllers may fail when the robot cannot reach contacts capable of rejecting a given disturbance. In this paper, we propose a search-based footstep planner which aims to maximize the probability of the robot successfully reaching the goal without falling as a result of a disturbance. The planner considers not only the poses of the planned contact sequence, but also alternative contacts near the planned contact sequence that can be used to recover from external disturbances. Although this additional consideration significantly increases the computation load, we train neural networks to efficiently predict multi-contact zero-step and one-step capturability, which allows the planner to generate robust contact sequences efficiently. Our results show that our approach generates footstep sequences that are more robust to external disturbances than a conventional footstep planner in four challenging scenarios.