Drive in Corridors: Enhancing the Safety of End-to-end Autonomous Driving via Corridor Learning and Planning

2025-04-10

AI Summary

End-to-end autonomous driving suffers from insufficient safety guarantees, primarily due to the absence of explicit behavioral constraints. To address this, we propose learning spatiotemporal collision-free corridors as an intermediate representation, explicitly embedding safety constraints into the end-to-end learning pipeline. Our method introduces a differentiable, learnable corridor prediction module that serves as an interpretable and constraint-aware bridge between perception and planning. By integrating corridor modeling, multi-task loss design, and differentiable nonlinear trajectory optimization, we jointly enhance both safety and driving performance. Evaluated on nuScenes, our approach reduces pedestrian and vehicle collision rates by 66.7%, decreases curb collision rate by 46.5%, and significantly improves closed-loop success rate—demonstrating effective safety–performance co-optimization within an end-to-end framework.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Natural Language Processing: Safety and Robustness
• Search and Optimization: Learning to Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Security and Privacy: Data transparency and provenance

Abstract

Safety remains one of the most critical challenges in autonomous driving systems. In recent years, the end-to-end driving has shown great promise in advancing vehicle autonomy in a scalable manner. However, existing approaches often face safety risks due to the lack of explicit behavior constraints. To address this issue, we uncover a new paradigm by introducing the corridor as the intermediate representation. Widely adopted in robotics planning, the corridors represents spatio-temporal obstacle-free zones for the vehicle to traverse. To ensure accurate corridor prediction in diverse traffic scenarios, we develop a comprehensive learning pipeline including data annotation, architecture refinement and loss formulation. The predicted corridor is further integrated as the constraint in a trajectory optimization process. By extending the differentiability of the optimization, we enable the optimized trajectory to be seamlessly trained within the end-to-end learning framework, improving both safety and interpretability. Experimental results on the nuScenes dataset demonstrate state-of-the-art performance of our approach, showing a 66.7% reduction in collisions with agents and a 46.5% reduction with curbs, significantly enhancing the safety of end-to-end driving. Additionally, incorporating the corridor contributes to higher success rates in closed-loop evaluations.

Problem

• Enhancing safety in end-to-end autonomous driving via corridor learning
• Reducing collisions using spatio-temporal obstacle-free corridor prediction
• Integrating corridor constraints into trajectory optimization for improved interpretability

Innovation

• Introducing corridor as intermediate representation
• Developing comprehensive learning pipeline for corridors
• Integrating corridor constraints in trajectory optimization