Safe Distributed Control of Multi-Robot Systems with Communication Delays

2024-02-14

IEEE Transactions on Vehicular Technology

AI Summary

This work addresses safety-critical cooperative control of multi-robot systems under significant communication delays—e.g., in underwater, cave, and deep-space environments. We propose a delay-aware distributed safety control framework. Our key contributions are: (1) the first formal definition and construction of Distributed Control Barrier Functions (DCBFs), enabling provably correct distributed verification and enforcement of safety constraints; and (2) an end-to-end trainable framework integrating state predictors, Graph Neural Networks (GNNs), and Age-of-Information (AoI) modeling to explicitly account for communication latency and data freshness. Evaluated on multi-robot collision avoidance, our method achieves over 40% higher safety success rate compared to delay-agnostic baselines, significantly improving system safety and robustness under realistic bandwidth- and latency-constrained communication conditions.

Technology Category

• Multiagent Systems: Agent Communication
• Natural Language Processing: Safety and Robustness
• Intelligent Robots: Multi-Robot Systems

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: Graph neural networks and deep learning approaches for Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Safe operation of multi-robot systems is critical, especially in communication-degraded environments such as underwater for seabed mapping, underground caves for navigation, and in extraterrestrial missions for assembly and construction. We address safety of networked autonomous systems where the information exchanged between robots incurs communication delays. We formalize a notion of distributed control barrier function for multi-robot systems, a safety certificate amenable to a distributed implementation, which provides formal ground to using graph neural networks to learn safe distributed controllers. Further, we observe that learning a distributed controller ignoring delays can severely degrade safety. We finally propose a predictor-based framework to train a safe distributed controller under communication delays, where the current state of nearby robots is predicted from received data and age-of-information. Numerical experiments on multi-robot collision avoidance show that our predictor-based approach can significantly improve the safety of a learned distributed controller under communication delays. A video abstract is available at https://youtu.be/Hcu1Ri32Spk.

Problem

• Ensuring safety in multi-robot systems with communication delays.
• Developing distributed control barrier functions for safe operations.
• Proposing a predictor-based framework to enhance controller safety.

Innovation

• Distributed control barrier function for safety
• Graph neural networks for distributed controllers
• Predictor-based framework for communication delays