Monte Carlo Tree Search with Branch and Bound for Multi-Robot Task Allocation

Multi-robot teams are effective in a variety of task allocation domains such as warehouse automation and surveillance. Robots in such domains have to perform tasks at given locations and specific times. Tasks have to be allocated to optimize given team objectives, such as minimizing the total distance traveled. We propose an efficient, satisficing and centralized Monte Carlo Tree Search (MCTS) based algorithm which exploits the branch and bound paradigm with a novel search parallelization method to solve the multi-robot task allocation problem with spatial, temporal and other side constraints. Unlike previous heuristics proposed for this problem, our approach maintains asymptotic convergence guarantees of MCTS and it has efficient anytime behavior. It finds near-optimal solutions for non-trivial problems in the Solomon data sets in an hour.