Optimal Task Assignment and Path Planning using Conflict-Based Search with Precedence and Temporal Constraints

Yu Quan Chong,Jiaoyang Li,Katia Sycara

2024-04-22

Abstract:The Multi-Agent Path Finding (MAPF) problem entails finding collision-free paths for a set of agents, guiding them from their start to goal locations. However, MAPF does not account for several practical task-related constraints. For example, agents may need to perform actions at goal locations with specific execution times, adhering to predetermined orders and timeframes. Moreover, goal assignments may not be predefined for agents, and the optimization objective may lack an explicit definition. To incorporate task assignment, path planning, and a user-defined objective into a coherent framework, this paper examines the Task Assignment and Path Finding with Precedence and Temporal Constraints (TAPF-PTC) problem. We augment Conflict-Based Search (CBS) to simultaneously generate task assignments and collision-free paths that adhere to precedence and temporal constraints, maximizing an objective quantified by the return from a user-defined reward function in reinforcement learning (RL). Experimentally, we demonstrate that our algorithm, CBS-TA-PTC, can solve highly challenging bomb-defusing tasks with precedence and temporal constraints efficiently relative to MARL and adapted Target Assignment and Path Finding (TAPF) methods.

Artificial Intelligence,Multiagent Systems

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

This paper aims to address the issues of Precedence and Temporal Constraints (PTC) in Task Assignment and Path Finding (TAPF) for multi-agent systems. Specifically, existing Multi-Agent Path Finding (MAPF) algorithms typically do not consider practical task-related constraints, such as specific execution times and order requirements, when generating collision-free paths. Additionally, goal assignments may not be predefined, and optimization objectives may lack clear definitions.

The paper proposes a new framework that combines Conflict-Based Search (CBS) methods to simultaneously generate task assignments and collision-free paths that satisfy precedence and temporal constraints, while maximizing the value returned by a user-defined reward function. Experimental results show that the proposed algorithm, CBS-TA-PTC, can efficiently solve complex bomb disposal tasks with precedence and temporal constraints, outperforming Multi-Agent Reinforcement Learning (MARL) and improved TAPF methods.