SADCHER: Scheduling using Attention-based Dynamic Coalitions of Heterogeneous Robots in Real-Time

2025-10-16

AI Summary

This work addresses real-time task allocation for heterogeneous multi-robot systems operating in dynamic, spatiotemporally non-uniform environments. Methodologically, we propose the first end-to-end scheduling framework that jointly integrates dynamic coalition formation and task priority constraints. Our approach employs a hybrid imitation learning architecture combining graph attention networks and Transformers to jointly encode spatiotemporal features—including robot positions, task durations, and residual processing times—while incorporating relaxed bipartite matching to ensure solution feasibility and computational efficiency. Innovatively, we unify adaptive teaming mechanisms with hard real-time requirements, substantially enhancing cross-environment generalization. Experiments demonstrate millisecond-scale inference latency on benchmark instances involving hundreds of tasks and robots, with consistent superiority over both learning-based and heuristic baselines on unseen stochastic task instances. To foster reproducibility and further research, we publicly release a large-scale benchmark dataset comprising 250,000 optimal scheduling solutions.

Technology Category

• Intelligent Robots: Multi-Robot Systems
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Humans and AI: Planning and Decision Support for Human-Machine Teams

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We present Sadcher, a real-time task assignment framework for heterogeneous multi-robot teams that incorporates dynamic coalition formation and task precedence constraints. Sadcher is trained through Imitation Learning and combines graph attention and transformers to predict assignment rewards between robots and tasks. Based on the predicted rewards, a relaxed bipartite matching step generates high-quality schedules with feasibility guarantees. We explicitly model robot and task positions, task durations, and robots'remaining processing times, enabling advanced temporal and spatial reasoning and generalization to environments with different spatiotemporal distributions compared to training. Trained on optimally solved small-scale instances, our method can scale to larger task sets and team sizes. Sadcher outperforms other learning-based and heuristic baselines on randomized, unseen problems for small and medium-sized teams with computation times suitable for real-time operation. We also explore sampling-based variants and evaluate scalability across robot and task counts. In addition, we release our dataset of 250,000 optimal schedules: https://autonomousrobots.nl/paper_websites/sadcher_MRTA/

Problem

• Dynamic coalition formation for heterogeneous robot task assignment
• Real-time scheduling with spatiotemporal reasoning and generalization
• Scalable learning from optimal small instances to larger problems

Innovation

• Dynamic coalition formation with task precedence constraints
• Graph attention and transformers predict assignment rewards
• Relaxed bipartite matching ensures feasible real-time schedules