SPARC: Spatial-Aware Path Planning via Attentive Robot Communication

2026-03-03

AI Summary

This work addresses the inefficiency of existing learning-based communication methods in decentralized multi-robot path planning, which often neglect spatial proximity among neighbors, leading to poor coordination in high-density scenarios. To overcome this limitation, the authors propose a Relation-enhanced Multi-head Attention (RMHA) mechanism integrated within the MAPPO framework. RMHA explicitly incorporates Manhattan distance to dynamically weight neighboring agents’ messages and combines a distance-constrained attention mask with GRU-gated message fusion, enabling end-to-end training. Notably, this approach is the first to embed spatial distance directly into the communication attention mechanism. Evaluated on a 40×40 grid with 30% obstacle density, the method achieves zero-shot generalization from 8 trained robots to 128 test robots, attaining a task success rate of approximately 75%—an improvement of over 25 percentage points compared to the best baseline.

Technology Category

• Multiagent Systems: Agent Communication
• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Efficient communication is critical for decentralized Multi-Robot Path Planning (MRPP), yet existing learned communication methods treat all neighboring robots equally regardless of their spatial proximity, leading to diluted attention in congested regions where coordination matters most. We propose Relation enhanced Multi Head Attention (RMHA), a communication mechanism that explicitly embeds pairwise Manhattan distances into the attention weight computation, enabling each robot to dynamically prioritize messages from spatially relevant neighbors. Combined with a distance-constrained attention mask and GRU gated message fusion, RMHA integrates seamlessly with MAPPO for stable end-to-end training. In zero-shot generalization from 8 training robots to 128 test robots on 40x40 grids, RMHA achieves approximately 75 percent success rate at 30 percent obstacle density outperforming the best baseline by over 25 percentage points. Ablation studies confirm that distance-relation encoding is the key contributor to success rate improvement in high-density environments. Index Terms-Multi-robot path planning, graph attention mechanism, multi-head attention, communication optimization, cooperative decision-making

Problem

• Multi-robot path planning
• communication optimization
• spatial awareness
• cooperative decision-making
• graph attention mechanism

Innovation

• Relation enhanced Multi Head Attention
• spatial-aware communication
• distance-constrained attention
• multi-robot path planning
• zero-shot generalization