webmaster: Sven Koenig

Learn all about Multi-Agent Path Finding (MAPF)


M. Liu, H. Ma, J. Li and S. Koenig. Task and Path Planning for Multi-Agent Pickup and Delivery. In Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS), pages (in print), 2019.

Abstract: We study the offline Multi-Agent Pickup-and-Delivery (MAPD) problem, where a team of agents has to execute a batch of tasks with release times in a known environment. To execute a task, an agent has to move first from its current location to the pickup location of the task and then to the delivery location of the task. The MAPD problem is to assign tasks to agents and plan collision-free paths for them to execute their tasks. Online MAPD algorithms can be applied to the offline MAPD problem, but do not utilize all of the available information and may thus not be effective. Therefore, we present two novel offline MAPD algorithms that improve a state-of-the-art online MAPD algorithm with respect to task planning, path planning, and deadlock avoidance for the offline MAPD problem. Our MAPD algorithms first compute one task sequence for each agent by solving a special traveling salesman problem and then plan paths according to these task sequences. We also introduce an effective deadlock avoidance method, called 'reserving dummy paths.' Theoretically, our MAPD algorithms are complete for well-formed MAPD instances, a realistic subclass of all MAPD instances. Experimentally, they produce solutions of smaller makespans and scale better than the online MAPD algorithm in simulated warehouses with hundreds of robots and thousands of tasks.

Download the paper in pdf.

(last updated in 2022)