Multi-Agent Reinforcement Learning for Safe Autonomous Driving Under Pedestrian Behavioral Uncertainty
A new study explores the use of multi-agent reinforcement learning (MARL) to improve the safety of autonomous vehicles in unpredictable pedestrian environments. The research indicates that co-training self-driving cars with pedestrians can significantly reduce collision rates compared to traditional methods. The findings suggest that incorporating pedestrian behavioral uncertainty leads to more realistic and effective safety assessments.
- ▪The study introduces a MARL environment where a self-driving car and 12 pedestrians are co-trained.
- ▪In evaluations, the co-trained self-driving car achieved a 78% goal completion rate with a 14% collision rate.
- ▪Jaywalking accounted for 13% of crossing events but was linked to 62% of collisions.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Machine Learning arXiv:2605.20255 (cs) [Submitted on 18 May 2026] Title:Multi-Agent Reinforcement Learning for Safe Autonomous Driving Under Pedestrian Behavioral Uncertainty Authors:Prakash Aryan, Kaushik Raghupathruni, Timo Kehrer, Sebastiano Panichella View a PDF of the paper titled Multi-Agent Reinforcement Learning for Safe Autonomous Driving Under Pedestrian Behavioral Uncertainty, by Prakash Aryan and 3 other authors View PDF HTML (experimental) Abstract:Simulation-based testing of self-driving cars (SDCs) typically relies on scripted or simplified pedestrian models that do not capture the heterogeneity and uncertainty of real human crossing behavior.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
