Abstract: We study the problem of multi-agent path finding with earliness-tardiness costs (MAPF-ET), which encourages agents to finish their tasks just-in-time, neither too early nor too late. Given ...