Multi-robot adversarial patrolling strategies via lattice paths

Jan Burmann, Jie Zhang

Research output: Contribution to conferencePaperpeer-review

Abstract

In full-knowledge multi-robot adversarial patrolling, a group of robots have to detect an adversary who knows the robots' strategy. The adversary can easily take advantage of any deterministic patrolling strategy, which necessitates the employment of a randomised strategy. While the Markov decision process has been the dominant methodology in computing the penetration detection probabilities, we apply enumerative combinatorics to characterise the penetration detection probabilities. It allows us to provide the closed formulae of these probabilities and facilitates characterising optimal random defence strategies. Comparing to iteratively updating the Markov transition matrices, our methods significantly reduces the time and space complexity of solving the problem. We use this method to tackle four penetration configurations.
Original languageEnglish
Publication statusPublished - 17 Jul 2020
Event29th International Joint Conference on Artificial Intelligence -
Duration: 11 Jul 202017 Jul 2020

Conference

Conference29th International Joint Conference on Artificial Intelligence
Abbreviated titleIJCAI 2020
Period11/07/2017/07/20

Keywords

  • Robot Planning
  • Planning under Uncertainty
  • Theoretical Foundations of Planning

Fingerprint

Dive into the research topics of 'Multi-robot adversarial patrolling strategies via lattice paths'. Together they form a unique fingerprint.

Cite this