Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Hypergraph-based Coordinated Task Allocation and Socially-aware Navigation for Multi-Robot Systems

Weizheng Wang, Aniket Bera, Byung-Cheol Min

TL;DR

Hyper-SAMARL addresses the challenge of coordinating multiple robots in human-populated environments by integrating a hypergraph diffusion mechanism with spatial-temporal transformer features and MAPPO-based MARL to jointly optimize dynamic task allocation and socially-aware navigation. It models high-order interactions among robots, pedestrians, and POIs via a hypergraph, and uses nonlinear diffusion to adapt to real-time changes. The framework is trained under the centralized training and decentralized execution paradigm and evaluated in dynamic simulations, showing improvements in social navigation, task completion efficiency, and adaptability over baselines. The results demonstrate the practical potential of hypergraph-based, MARL-driven collaboration for safe and efficient multi-robot operations in public spaces.

Abstract

A team of multiple robots seamlessly and safely working in human-filled public environments requires adaptive task allocation and socially-aware navigation that account for dynamic human behavior. Current approaches struggle with highly dynamic pedestrian movement and the need for flexible task allocation. We propose Hyper-SAMARL, a hypergraph-based system for multi-robot task allocation and socially-aware navigation, leveraging multi-agent reinforcement learning (MARL). Hyper-SAMARL models the environmental dynamics between robots, humans, and points of interest (POIs) using a hypergraph, enabling adaptive task assignment and socially-compliant navigation through a hypergraph diffusion mechanism. Our framework, trained with MARL, effectively captures interactions between robots and humans, adapting tasks based on real-time changes in human activity. Experimental results demonstrate that Hyper-SAMARL outperforms baseline models in terms of social navigation, task completion efficiency, and adaptability in various simulated scenarios.

Hypergraph-based Coordinated Task Allocation and Socially-aware Navigation for Multi-Robot Systems

TL;DR

Hyper-SAMARL addresses the challenge of coordinating multiple robots in human-populated environments by integrating a hypergraph diffusion mechanism with spatial-temporal transformer features and MAPPO-based MARL to jointly optimize dynamic task allocation and socially-aware navigation. It models high-order interactions among robots, pedestrians, and POIs via a hypergraph, and uses nonlinear diffusion to adapt to real-time changes. The framework is trained under the centralized training and decentralized execution paradigm and evaluated in dynamic simulations, showing improvements in social navigation, task completion efficiency, and adaptability over baselines. The results demonstrate the practical potential of hypergraph-based, MARL-driven collaboration for safe and efficient multi-robot operations in public spaces.

Abstract

A team of multiple robots seamlessly and safely working in human-filled public environments requires adaptive task allocation and socially-aware navigation that account for dynamic human behavior. Current approaches struggle with highly dynamic pedestrian movement and the need for flexible task allocation. We propose Hyper-SAMARL, a hypergraph-based system for multi-robot task allocation and socially-aware navigation, leveraging multi-agent reinforcement learning (MARL). Hyper-SAMARL models the environmental dynamics between robots, humans, and points of interest (POIs) using a hypergraph, enabling adaptive task assignment and socially-compliant navigation through a hypergraph diffusion mechanism. Our framework, trained with MARL, effectively captures interactions between robots and humans, adapting tasks based on real-time changes in human activity. Experimental results demonstrate that Hyper-SAMARL outperforms baseline models in terms of social navigation, task completion efficiency, and adaptability in various simulated scenarios.
Paper Structure (17 sections, 12 equations, 3 figures, 1 table, 1 algorithm)

This paper contains 17 sections, 12 equations, 3 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Preliminary
  4. Hypergraph Dynamic Relational Reasoning
  5. Markov Decision Process Formulation
  6. Robots Task Allocation $\&$ Social Navigation Statement
  7. Methodology
  8. Spatial-Temporal Transformer
  9. Hypergraph Diffusion
  10. Multi-Agent Reinforcement Learning
  11. Experiments And Results
  12. Simulation Environment
  13. Baselines and Ablation Study
  14. Training Details
  15. Evaluation
  16. ...and 2 more sections

Figures (3)

  • Figure 1: An illustration of multi-robot task allocation and social navigation task and simulator: mobile robots are engaging in cooperative navigation toward POIs in a human-filled environment while adhering to social norms.
  • Figure 2: Hyper-SAMARL Architecture: First, task information related to points of interest (POIs) and robot observations are encoded by the positional embedding encoder of the transformer as spatial-temporal input. Next, the hypergraph for MR-TASN is initialized using attention-based vertex features and Euclidean-based hyperedge features. The hypergraph diffusion mechanism is then employed to propagate vertex features across the hypergraph structure, ensuring balanced hypergraph dynamics. Finally, the diffused hypergraph features are decoded by the robot policy to generate macro-actions (MA) and local actions (LA), which are trained using the MAPPO algorithm.
  • Figure 3: Learning curves of Hyper-SAMARL and other two ablation models with five different seeds.