CarlaNCAP: A Framework for Quantifying the Safety of Vulnerable Road Users in Infrastructure-Assisted Collective Perception Using EuroNCAP Scenarios

2025-12-12

AI Summary

Vulnerable road users (VRUs) in urban environments suffer from occlusion-induced perception blind spots, and ego-vehicle sensor–only systems achieve low accident avoidance rates (33%). Method: We propose CarlaNCAP—the first infrastructure-assisted collaborative perception quantification framework tailored for VRUs—and construct the first 11k-frame roadside safety-critical dataset aligned with Euro NCAP standards. Leveraging CARLA, we model multi-agent collaborative perception, optimize roadside sensor placement, and develop a spatiotemporal VRU risk model supported by statistical accident rate analysis. Contribution/Results: In safety-critical scenarios, roadside assistance achieves 100% accident avoidance—significantly outperforming ego-vehicle–only approaches. The open-sourced code and dataset enable standardized industry validation and advance the development of integrated vehicle–road–cloud safety evaluation systems.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Safety and Robustness

Application Category

• Security and Privacy: Large-scale security measurements
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

The growing number of road users has significantly increased the risk of accidents in recent years. Vulnerable Road Users (VRUs) are particularly at risk, especially in urban environments where they are often occluded by parked vehicles or buildings. Autonomous Driving (AD) and Collective Perception (CP) are promising solutions to mitigate these risks. In particular, infrastructure-assisted CP, where sensor units are mounted on infrastructure elements such as traffic lights or lamp posts, can help overcome perceptual limitations by providing enhanced points of view, which significantly reduces occlusions. To encourage decision makers to adopt this technology, comprehensive studies and datasets demonstrating safety improvements for VRUs are essential. In this paper, we propose a framework for evaluating the safety improvement by infrastructure-based CP specifically targeted at VRUs including a dataset with safety-critical EuroNCAP scenarios (CarlaNCAP) with 11k frames. Using this dataset, we conduct an in-depth simulation study and demonstrate that infrastructure-assisted CP can significantly reduce accident rates in safety-critical scenarios, achieving up to 100% accident avoidance compared to a vehicle equipped with sensors with only 33%. Code is available at https://github.com/ekut-es/carla_ncap

Problem

• Evaluates infrastructure-assisted Collective Perception safety for Vulnerable Road Users
• Addresses occlusion risks in urban environments using EuroNCAP scenarios
• Quantifies accident reduction through simulation with a dedicated dataset

Innovation

• Framework for evaluating infrastructure-assisted collective perception safety
• Dataset with EuroNCAP scenarios for VRU safety analysis
• Simulation showing up to 100% accident avoidance improvement