Congestion Mitigation Path Planning for Large-Scale Multi-Agent Navigation in Dense Environments

TL;DR

This paper introduces a congestion-aware path planning method for large-scale multi-agent navigation, embedding congestion costs into route optimization to improve efficiency and reduce local congestion in dense environments.

Contribution

It formulates the novel CMPP problem, develops scalable solvers including an exact and an approximate algorithm, and demonstrates significant performance improvements in multi-agent systems.

Findings

CMPP reduces local congestion in multi-agent navigation.

Augmenting collision-avoidance with CMPP improves throughput.

Scalable algorithms effectively handle large-scale instances.

Abstract

In high-density environments where numerous autonomous agents move simultaneously in a distributed manner, streamlining global flows to mitigate local congestion is crucial to maintain overall navigation efficiency. This paper introduces a novel path-planning problem, congestion mitigation path planning (CMPP), which embeds congestion directly into the cost function, defined by the usage of incoming edges along agents' paths. CMPP assigns a flow-based multiplicative penalty to each vertex of a sparse graph, which grows steeply where frequently-traversed paths intersect, capturing the intuition that congestion intensifies where many agents enter the same area from different directions. Minimizing the total cost yields a set of coarse-level, time-independent routes that autonomous agents can follow while applying their own local collision avoidance. We formulate the problem and develop two solvers: (i) an exact mixed-integer nonlinear programming solver for small instances, and (ii) a scalable two-layer search algorithm, A-CMTS, which quickly finds suboptimal solutions for large-scale instances and iteratively refines them toward the optimum. Empirical studies show that augmenting state-of-the-art collision-avoidance planners with CMPP significantly reduces local congestion and enhances system throughput in both discrete- and continuous-space scenarios. These results indicate that CMPP improves the performance of multi-agent systems in real-world applications such as logistics and autonomous-vehicle operations.