A Survey on Multi-agent Pathfinding Problem

PhD Qualifying Examination


Title: "A Survey on Multi-agent Pathfinding Problem"

by

Mr. Mingkai TANG


Abstract:

The multi-agent pathfinding (MAPF) problem aims to determine paths for multiple 
agents so that each reaches its goal without collisions. Recently, the MAPF 
problem has garnered significant attention within the research communities of 
Artificial Intelligence (AI), Multi-Agent Systems (MAS), and Robotics due to 
its numerous real-world applications, such as warehouse management, traffic 
control, aircraft scheduling, and virtual network embedding. However, solving 
the MAPF problem remains challenging due to its high computational complexity.

In this survey, we first present the formal definition of the classical MAPF 
problem. We then provide a systematic overview of current research on the 
classical MAPF problem and describe a commonly used benchmark set in this 
field. Additionally, we discuss works that relax some assumptions of the 
classical MAPF problem to make it more applicable to various scenarios. 
Finally, we suggest promising directions for future research and conclude the 
survey.


Date:                   Tuesday, 20 August 2024

Time:                   3:00pm - 5:00pm

Venue:                  Room 3494
                        Lifts 25/26

Committee Members:      Dr. Ling Pan (Supervisor)
                        Prof. Fangzhen Lin (Chairperson)
                        Prof. Qiong Luo
                        Dr. Yangqiu Song