Keywords: multi-robot planning, coordination, topological graphs
TL;DR: We plan coordinated multi-robot tactics through complex environments to minimize team visibility and maximize safety in an adversarial environment.
Abstract: Achieving unified multi-robot coordination and motion planning in complex, off-road or rugged urban environments is a challenging problem. In this paper, we present a hierarchical approach for minimizing visibility and maximizing safety with a multi-robot team navigating through a hazardous environment. In particular, our approach first segments the environment based on the visibility of the robot team from an adversarial observer’s perspective, creates a topological graph, and computes an optimal multi-robot plan on that graph using mixed-integer programming (MIP). This visibility information also informs the lower layers of the autonomy stack to achieve dynamically feasible multi-robot planning across the hierarchy. We then demonstrate our approach in simulation in an off-road environment with multiple vehicles and on hardware in a rugged urban environment with a single autonomous Clearpath Warthog. We also discuss future plans for hardware demonstrations of multi-vehicle operations in off-road environments and the incorporation of learning to compensate for additional environmental complexities.
Submission Number: 10
Loading