Keisuke Okumura

BIOgraphy

Dr. Keisuke Okumura is a postdoctoral scholar at the University of Cambridge, UK, and also a research scientist at the National Institute of Advanced Industrial Science and Technology, Japan. He received his Ph.D. in computer science from Tokyo Institute of Technology (TokyoTech) in 2023. His research interests are in designing intelligent collective behavior for swarm agents, particularly the development of multi-agent planning methods tailored to large-scale automation. He has been honored with several research awards, including the Best Student Paper Award at ICAPS 2022 and the Ph.D. Dissertation Award at TokyoTech in 2024.

Title: Pathfinding for 10k agents

Abstract

Multi-agent pathfinding (MAPF) is an attractive problem that can facilitate efficient large-scale automation. This talk dives deep into an important challenge for MAPF: can we build scalable algorithms, say tailored for 10k agents, while still having nice theoretical guarantees like completeness and optimality? Traditionally, these two aspects have been considered incompatible, given the inherent difficulties that guaranteed algorithms face in finding solutions amidst an increasing number of agents. Starting with a vanilla A* search, this talk outlines approaches to this quest.