Technology Category
Multiagent Systems: Distributed Problem SolvingPlanning, Routing, and Scheduling: Learning for Planning and SchedulingIntelligent Robots: Learning & Optimization for ROB
HIPPO-MAT: Decentralized Task Allocation Using GraphSAGE and Multi-Agent Deep Reinforcement Learning
🤖 AI Summary
This work addresses decentralized continuous task allocation for heterogeneous multi-agent systems (UAVs/UGVs) operating in dynamic environments, aiming for cost-optimal and collision-free collaborative execution. Methodologically, it introduces the first distributed reinforcement learning framework integrating GraphSAGE—a graph neural network—for agent interaction modeling with independent PPO (IPPO) for decentralized policy optimization; further coupled with an enhanced A* planner and SLAM for online obstacle avoidance and real-time responsiveness, all deployed on Jetson Nano via ESP-NOW within a ROS architecture. Experiments demonstrate a 92.5% task success rate, only 16.49% suboptimal versus centralized Hungarian assignment, scalability to 30 agents, and a per-step decision latency of 0.32 simulation steps—validated on a JetBot platform. The core contribution is the first lightweight, real-time task allocation framework that deeply unifies graph representation learning with decentralized policy optimization.
Technology Category
Application Category
📝 Abstract
This paper tackles decentralized continuous task allocation in heterogeneous multi-agent systems. We present a novel framework HIPPO-MAT that integrates graph neural networks (GNN) employing a GraphSAGE architecture to compute independent embeddings on each agent with an Independent Proximal Policy Optimization (IPPO) approach for multi-agent deep reinforcement learning. In our system, unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) share aggregated observation data via communication channels while independently processing these inputs to generate enriched state embeddings. This design enables dynamic, cost-optimal, conflict-aware task allocation in a 3D grid environment without the need for centralized coordination. A modified A* path planner is incorporated for efficient routing and collision avoidance. Simulation experiments demonstrate scalability with up to 30 agents and preliminary real-world validation on JetBot ROS AI Robots, each running its model on a Jetson Nano and communicating through an ESP-NOW protocol using ESP32-S3, which confirms the practical viability of the approach that incorporates simultaneous localization and mapping (SLAM). Experimental results revealed that our method achieves a high 92.5% conflict-free success rate, with only a 16.49% performance gap compared to the centralized Hungarian method, while outperforming the heuristic decentralized baseline based on greedy approach. Additionally, the framework exhibits scalability with up to 30 agents with allocation processing of 0.32 simulation step time and robustness in responding to dynamically generated tasks.
Problem
Research questions and friction points this paper is trying to address.
Decentralized task allocation in heterogeneous multi-agent systems.
Dynamic, cost-optimal, conflict-aware task allocation in 3D grid.
Scalable and robust framework for real-world multi-agent applications.
Innovation
Methods, ideas, or system contributions that make the work stand out.
GraphSAGE for decentralized agent embeddings
IPPO for multi-agent reinforcement learning
Modified A* planner for efficient routing
🔎 Similar Papers
Multi-Robot Informative Path Planning for Efficient Target Mapping using Deep Reinforcement Learning
2024-09-25arXiv.orgCitations: 0
