WinkTPG: An Execution Framework for Multi-Agent Path Finding Using Temporal Reasoning

2025-08-02

AI Summary

To address the dynamical mismatch problem—where ideal paths generated by large-scale Multi-Agent Path Finding (MAPF) solvers are infeasible for direct execution—we propose kTPG and WinkTPG, the first frameworks to integrate windowed Temporal Plan Graphs (TPGs) into multi-agent cooperative trajectory generation. Our approach unifies modeling of kinematic constraints, spatiotemporal collision avoidance, and execution uncertainty through a sliding-window mechanism, velocity optimization, and incremental replanning. It ensures real-time performance—generating trajectories for 1,000 agents within one second—while significantly improving trajectory feasibility and execution robustness. Experimental results demonstrate up to a 51.7% improvement in solution quality over state-of-the-art methods, substantially enhancing the practical deployability of large-scale multi-agent systems.

Technology Category

• Multiagent Systems: Multiagent Planning
• Planning, Routing, and Scheduling: Temporal Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Planning collision-free paths for a large group of agents is a challenging problem with numerous real-world applications. While recent advances in Multi-Agent Path Finding (MAPF) have shown promising progress, standard MAPF algorithms rely on simplified kinodynamic models, preventing agents from directly following the generated MAPF plan. To bridge this gap, we propose kinodynamic Temporal Plan Graph Planning (kTPG), a multi-agent speed optimization algorithm that efficiently refines a MAPF plan into a kinodynamically feasible plan while accounting for uncertainties and preserving collision-freeness. Building on kTPG, we propose Windowed kTPG (WinkTPG), a MAPF execution framework that incrementally refines MAPF plans using a window-based mechanism, dynamically incorporating agent information during execution to reduce uncertainty. Experiments show that WinkTPG can generate speed profiles for up to 1,000 agents in 1 second and improves solution quality by up to 51.7% over existing MAPF execution methods.

Problem

• Planning collision-free paths for multiple agents efficiently
• Refining MAPF plans into kinodynamically feasible solutions
• Reducing execution uncertainty with dynamic window-based updates

Innovation

• Uses kinodynamic Temporal Plan Graph Planning
• Incorporates window-based incremental refinement
• Handles uncertainties while ensuring collision-freeness