The ideal vulnerable road user - a study of parameters affecting VRU detection

Area of Science:

• Automotive Engineering
• Road Safety Research
• Human-Machine Interaction

Background:

• Advanced Driving Assistance Systems (ADAS) aim to reduce road collisions with Vulnerable Road Users (VRUs).
• Detecting diverse pedestrian appearances presents a significant technical challenge for ADAS.
• Accurate detection is crucial for effective collision mitigation measures.

Purpose of the Study:

• To evaluate how various parameters influence the pedestrian detection capabilities of ADAS.
• To benchmark system performance under diverse and challenging real-world-like conditions through track testing.

Main Methods:

• Utilized standardized adult (EPTa) and child (EPTc) pedestrian targets for track testing.
• Followed Euro NCAP AEB-VRU protocols, conducting over 1,000 tests with 24 configurations across 13 vehicles (2019-2021 models).
• Varied environmental conditions (night, snow) and pedestrian factors (clothing, accessories, obstructions, multiple targets) to simulate urban scenarios.

Main Results:

• Parameter variations significantly impacted vehicle safety performance, with some ADAS more affected than others by weather and urban-like scenarios.
• Performance varied notably across different vehicle models and test conditions.
• Findings highlight the influence of environmental and pedestrian-specific factors on ADAS detection.

Conclusions:

• Canadian vehicles face diverse road conditions and pedestrian profiles, necessitating robust ADAS.
• Tested vehicles showed varied effectiveness in mitigating pedestrian collisions.
• Research underscores the safety risks posed by adverse weather and diverse Vulnerable Road User types.