Robust and Safe Multi-Agent Reinforcement Learning with Communication for Autonomous Vehicles: From Simulation to Hardware
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910213424021504 |
|---|---|
| author | Smith, Keshawn Zhang, Zhili Ahmad, H M Sabbir Sabouni, Ehsan Mondal, Mainak Han, Song Li, Wenchao Miao, Fei |
| author_facet | Smith, Keshawn Zhang, Zhili Ahmad, H M Sabbir Sabouni, Ehsan Mondal, Mainak Han, Song Li, Wenchao Miao, Fei |
| contents | Deep multi-agent reinforcement learning (MARL) has been demonstrated effectively in simulations for multi-robot problems. For autonomous vehicles, the development of vehicle-to-vehicle (V2V) communication technologies provide opportunities to further enhance system safety. However, zero-shot transfer of simulator-trained MARL policies to dynamic hardware systems remains challenging, and how to leverage communication and shared information for MARL has limited demonstrations on hardware. This problem is challenged by discrepancies between simulated and physical states, system state and model uncertainties, practical shared information design, and the need for safety guarantees in both simulation and hardware. This paper designs RSR-RSMARL, a novel Robust and Safe MARL framework that supports Real-Sim-Real (RSR) policy adaptation for multi-agent systems with communication among agents, with both simulation and hardware demonstrations. RSR-RSMARL leverages state (includes shared state information among agents) and action representations considering real system complexities for MARL formulation. The MARL policy is trained with robust MARL algorithm to enable zero-shot transfer to hardware considering the sim-to-real gap. A safety shield module using Control Barrier Functions (CBFs) provides safety guarantee for each individual agent. Experimental results on 1/10th-scale autonomous vehicles with V2V communication demonstrate the ability of RSR-RSMARL framework to enhance driving safety and coordination across multiple configurations. These findings emphasize the importance of jointly designing robust policy representations and modular safety architectures to enable scalable, generalizable RSR transfer in multi-agent autonomy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_00982 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Robust and Safe Multi-Agent Reinforcement Learning with Communication for Autonomous Vehicles: From Simulation to Hardware Smith, Keshawn Zhang, Zhili Ahmad, H M Sabbir Sabouni, Ehsan Mondal, Mainak Han, Song Li, Wenchao Miao, Fei Robotics Multiagent Systems Deep multi-agent reinforcement learning (MARL) has been demonstrated effectively in simulations for multi-robot problems. For autonomous vehicles, the development of vehicle-to-vehicle (V2V) communication technologies provide opportunities to further enhance system safety. However, zero-shot transfer of simulator-trained MARL policies to dynamic hardware systems remains challenging, and how to leverage communication and shared information for MARL has limited demonstrations on hardware. This problem is challenged by discrepancies between simulated and physical states, system state and model uncertainties, practical shared information design, and the need for safety guarantees in both simulation and hardware. This paper designs RSR-RSMARL, a novel Robust and Safe MARL framework that supports Real-Sim-Real (RSR) policy adaptation for multi-agent systems with communication among agents, with both simulation and hardware demonstrations. RSR-RSMARL leverages state (includes shared state information among agents) and action representations considering real system complexities for MARL formulation. The MARL policy is trained with robust MARL algorithm to enable zero-shot transfer to hardware considering the sim-to-real gap. A safety shield module using Control Barrier Functions (CBFs) provides safety guarantee for each individual agent. Experimental results on 1/10th-scale autonomous vehicles with V2V communication demonstrate the ability of RSR-RSMARL framework to enhance driving safety and coordination across multiple configurations. These findings emphasize the importance of jointly designing robust policy representations and modular safety architectures to enable scalable, generalizable RSR transfer in multi-agent autonomy. |
| title | Robust and Safe Multi-Agent Reinforcement Learning with Communication for Autonomous Vehicles: From Simulation to Hardware |
| topic | Robotics Multiagent Systems |
| url | https://arxiv.org/abs/2506.00982 |