Dynamic police patrol scheduling with multi-agent reinforcement learning

Effective police patrol scheduling is essential in projecting police presence and ensuring readiness in responding to unexpected events in urban environments. However, scheduling patrols can be a challenging task as it requires balancing between two conflicting objectives namely projecting presence...

Full description

Saved in:
Bibliographic Details
Main Authors: WONG, Songhan, JOE, Waldy, LAU, Hoong Chuin
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2023
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/8347
https://ink.library.smu.edu.sg/context/sis_research/article/9350/viewcontent/Dynamic.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
Description
Summary:Effective police patrol scheduling is essential in projecting police presence and ensuring readiness in responding to unexpected events in urban environments. However, scheduling patrols can be a challenging task as it requires balancing between two conflicting objectives namely projecting presence (proactive patrol) and incident response (reactive patrol). This task is made even more challenging with the fact that patrol schedules do not remain static as occurrences of dynamic incidents can disrupt the existing schedules. In this paper, we propose a solution to this problem using Multi-Agent Reinforcement Learning (MARL) to address the Dynamic Bi-objective Police Patrol Dispatching and Rescheduling Problem (DPRP). Our solution utilizes an Asynchronous Proximal Policy Optimization-based (APPO) actor-critic method that learns a policy to determine a set of prescribed dispatch rules to dynamically reschedule existing patrol plans. The proposed solution not only reduces computational time required for training, but also improves the solution quality in comparison to an existing RL-based approach that relies on heuristic solver.