Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Lightweight Traffic Map for Efficient Anytime LaCAM*

Bojie Shen, Yue Zhang, Zhe Chen, Daniel Harabor

TL;DR

This work proposes a new approach that leverages LaCAM*'s ability to construct a dynamic, lightweight traffic map during its search that achieves higher solution quality than state-of-the-art guidance-path approaches across two MAPF variants.

Abstract

Multi-Agent Path Finding (MAPF) aims to compute collision-free paths for multiple agents and has a wide range of practical applications. LaCAM*, an anytime configuration-based solver, currently represents the state of the art. Recent work has explored the use of guidance paths to steer LaCAM* toward configurations that avoid traffic congestion, thereby improving solution quality. However, existing approaches rely on Frank-Wolfe-style optimization that repeatedly invokes single-agent search before executing LaCAM*, resulting in substantial computational overhead for large-scale problems. Moreover, the guidance path is static and primarily beneficial for finding the first solution in LaCAM*. To address these limitations, we propose a new approach that leverages LaCAM*'s ability to construct a dynamic, lightweight traffic map during its search. Experimental results demonstrate that our method achieves higher solution quality than state-of-the-art guidance-path approaches across two MAPF variants.

A Lightweight Traffic Map for Efficient Anytime LaCAM*

TL;DR

This work proposes a new approach that leverages LaCAM*'s ability to construct a dynamic, lightweight traffic map during its search that achieves higher solution quality than state-of-the-art guidance-path approaches across two MAPF variants.

Abstract

Multi-Agent Path Finding (MAPF) aims to compute collision-free paths for multiple agents and has a wide range of practical applications. LaCAM*, an anytime configuration-based solver, currently represents the state of the art. Recent work has explored the use of guidance paths to steer LaCAM* toward configurations that avoid traffic congestion, thereby improving solution quality. However, existing approaches rely on Frank-Wolfe-style optimization that repeatedly invokes single-agent search before executing LaCAM*, resulting in substantial computational overhead for large-scale problems. Moreover, the guidance path is static and primarily beneficial for finding the first solution in LaCAM*. To address these limitations, we propose a new approach that leverages LaCAM*'s ability to construct a dynamic, lightweight traffic map during its search. Experimental results demonstrate that our method achieves higher solution quality than state-of-the-art guidance-path approaches across two MAPF variants.
Paper Structure (20 sections, 1 equation, 4 figures)

This paper contains 20 sections, 1 equation, 4 figures.

Table of Contents

  1. Introduction
  2. Problem Definition
  3. Sum-of-loss objective.
  4. Related Works
  5. PIBT and LaCAM
  6. Priority Inheritance and Backtracking (PIBT)
  7. Lazy Constraint Addition MAPF (LaCAM)
  8. Traffic Optimisation Using Guide Paths
  9. Our Method
  10. Updating the Lightweight Traffic Map
  11. Modifying LaCAM* with LTM
  12. Modifying LaCAM* with frequent restarts
  13. Putting it Together
  14. One-shot MAPF
  15. Planning and Execution MAPF
  16. ...and 5 more sections

Figures (4)

  • Figure 1: Comparison of solution quality on eight grid-based MAPF benchmarks under the one-shot setting. Each plot shows the sum of loss ratio as a function of the number of agents for LaCAM*, LaCAM*+TO, LaCAM*+SUO, and the proposed LaCAM*+LTM, averaged over 25 random instances with a 30-second time limit. Lower values indicate better solution quality.
  • Figure 2: Visualization of the Lightweight Traffic Map (LTM) on the random-32-32-20 map with 400 agents. Circles denote start locations and squares denote goal locations. Edge colors indicate normalized traffic intensity. The LTM is shown after early iterations (left) and after 30 seconds of search (right).
  • Figure 3: Coverage plot over a 30-second runtime for 1000 agents on the random-64-64-20.map, comparing LaCAM, LaCAM+TO, LaCAM*+SUO, and the proposed LaCAM*+LTM.
  • Figure 4: Planning-and-execution MAPF, each figure reports the sum of loss ratio for LaCAM*, LaCAM*+TO, LaCAM*+SUO, and the proposed LaCAM+LTM. Results are shown for different execution times (E = 0.1 s and 0.5 s) and commitment steps of 5, 10, and 20 steps, respectively. Values are averaged over instances; error bars indicate standard deviation. Lower values indicate better solution quality.

Theorems & Definitions (1)

  • Definition 1