Multi-Robot Coordinated Motion Planning at Intersections using Lattice-Guided dRRT*
Victor Parque, Tomoyuki Miyashita · 2023
Coordinated multi-robot motion planning at intersections is essential to enable efficient and safe mobility in roads, factories, and warehouse facilities. In recent years, the discrete RRT-based algorithms have shown the potential to enable collision-free trajectories for multi-robot configurations with reasonable computational time and solution quality; however, constructing initial local configuration spaces by Probabilistic Roadmaps (PRMs) is computationally expensive. In this paper, we study the feasibility of using lattice-based roadmap configurations and gradient-free optimization heuristics to compute collision-free multi-robot coordinated trajectories at intersections. Our computational experiments with diverse navigation scenarios have shown the merit and feasibility to compute collision-free trajectories within 500 function evaluations. Our results have the potential to render new gradient-free graphbased algorithms for multi-robot coordinated planning and navigation.