CooperDrive: Enhancing Driving Decisions Through Cooperative Perception

2026-04-15

AI Summary

This study addresses the challenges of perception degradation in occluded and non-line-of-sight (NLOS) scenarios, which can lead to delayed reactions and increased collision risk for autonomous vehicles. The authors propose a lightweight inter-vehicle cooperative perception framework that enhances situational awareness without requiring additional heavy-weight encoders. By sharing object-level information and reusing bird’s-eye-view (BEV) features while preserving each vehicle’s native perception and planning pipeline, the method enables efficient pose estimation and environmental modeling. Integrated with low-latency communication and trajectory planning, the system shifts driving behavior from reactive to proactive. Real-world closed-loop experiments demonstrate significant improvements in reaction lead time, minimum time-to-collision (TTC), and braking margin at highly occluded NLOS intersections, achieving these gains with only 90 kbps communication bandwidth and an end-to-end average latency of 89 ms.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Vision for Robotics & Autonomous Driving
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• 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

Abstract

Autonomous vehicles equipped with robust onboard perception, localization, and planning still face limitations in occlusion and non-line-of-sight (NLOS) scenarios, where delayed reactions can increase collision risk. We propose CooperDrive, a cooperative perception framework that augments situational awareness and enables earlier, safer driving decisions. CooperDrive offers two key advantages: (i) each vehicle retains its native perception, localization, and planning stack, and (ii) a lightweight object-level sharing and fusion strategy bridges perception and planning. Specifically, CooperDrive reuses detector Bird's-Eye View (BEV) features to estimate accurate vehicle poses without additional heavy encoders, thereby reconstructing BEV representations and feeding the planner with low latency. On the planning side, CooperDrive leverages the expanded object set to anticipate potential conflicts earlier and adjust speed and trajectory proactively, thereby transforming reactive behaviors into predictive and safer driving decisions. Real-world closed-loop tests at occlusion-heavy NLOS intersections demonstrate that CooperDrive increases reaction lead time, minimum time-to-collision (TTC), and stopping margin, while requiring only 90 kbps bandwidth and maintaining an average end-to-end latency of 89 ms.

Problem

• occlusion
• non-line-of-sight
• autonomous driving
• collision risk
• situational awareness

Innovation

• cooperative perception
• Bird's-Eye View (BEV)
• object-level fusion
• predictive planning
• low-latency communication