Spatiotemporal Elastic Bands for Motion Planning in Highly Dynamic Environments

Robust motion planning in highly dynamic environments affected by challenging conditions remains an important task for autonomous robots, and an open problem for the robotics community. This paper proposes significant extensions to the elastic band method that gives more robustness to uncertainty in state and tracking performance, and a way to avoid fast-moving obstacles that may move multiple times faster than the vehicle in an efficient and non-conservative way. Particularly, we temporally enhance the algorithm, address future collisions spatiotemporally with continuous guarantees, and adapt the required safety clearance dynamically to address disturbances, control errors, and uncertainty. To validate the proposed method, results from a simulation study are presented, demonstrating the ability to safely plan trajectories in dynamic environments.The motion planner is lightweight and remarkably computationally efficient, with replanning orders of magnitudes faster than real-time needs by reaching and surpassing 1000Hz. Copyright (C) 2024 The Authors. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)