This paper presents a reactive planning system that allows a Cassie-series bipedal robot to avoid multiple nonoverlapping obstacles via a single, continuously differentiable control barrier function (CBF). The overall system detects an individual obstacle via a height map derived from a LiDAR point cloud and computes an elliptical outer approximation, which is then turned into a CBF. The QP-CLF-CBF formalism developed by Ames et al. is applied to ensure that safe trajectories are generated. Liveness is ensured by an analysis of induced equilibrium points that are distinct from the goal state. Safe planning in environments with multiple obstacles is demonstrated both in simulation and experimentally on the Cassie biped.
THIS IS AN INITIAL DRAFTWhile the paper is not yet polished, it allows the cofirst authors to highlight their research skills while they are seeking a PhD position. The full autonomy videos are upload to our YouTube channel and the video for this particular paper can be viewed here. This draft has been approved by Huang and Grizzle.