Multi-Robot Spatio-Temporal Path Planning With Robot Contour Constraints

Multi-robot spatio-temporal path planning with robot contour constraints is proposed to solve the problem of MAPF solvers' inability to solve real-world MAPF problems. The top level of the algorithm performs a best-first search in a conflict tree whose nodes are a solution to the given MAPF problem. The paper also proposes a spatio-temporal A* improved algorithm as a bottom layer algorithm to find a route for individual robots. The route planned for the robot consists of steps. Each step contains the spatial and temporal information of the robot. A graph containing constraints is set before running the algorithm. The experiment is done for the algorithm with the conflict-based search (CBS) algorithm in simulation and real-world scenarios. The results show that the success rate and the correct rate of the algorithm are higher than the CBS algorithm, which proves the applicability and effectiveness of this algorithm in large-scale multi-robot scenarios in the real world.