Optimal and Bounded-Suboptimal Multi-Goal Task Assignment and Path Finding

TL;DR

This paper formalizes the multi-goal task assignment and path finding problem, proves its NP-hardness, and introduces algorithms for optimal and bounded-suboptimal solutions, with experimental comparisons across benchmarks.

Contribution

It introduces the MG-TAPF problem, proves its NP-hardness, and develops algorithms for optimal and suboptimal solutions with experimental validation.

Findings

MG-TAPF is NP-hard to solve optimally.

Algorithms for optimal and bounded-suboptimal solutions are proposed.

Experimental results compare algorithm performance across benchmarks.

Abstract

We formalize and study the multi-goal task assignment and path finding (MG-TAPF) problem from theoretical and algorithmic perspectives. The MG-TAPF problem is to compute an assignment of tasks to agents, where each task consists of a sequence of goal locations, and collision-free paths for the agents that visit all goal locations of their assigned tasks in sequence. Theoretically, we prove that the MG-TAPF problem is NP-hard to solve optimally. We present algorithms that build upon algorithmic techniques for the multi-agent path finding problem and solve the MG-TAPF problem optimally and bounded-suboptimally. We experimentally compare these algorithms on a variety of different benchmark domains.