FURI | Summer 2022

Driving Safety Performance Assessment Metrics for ADS-Equipped Vehicles and Scenarios Testing

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

One of the biggest questions facing the safe deployment and commercialization of automated driving systems (ADS)-equipped vehicles ​(AVs) today is “What level of driving safety performance is required compared to that of a human-driven vehicle?” A methodology to systematically compare and evaluate the driving safety performance of AVs and human-driven vehicles is needed to provide an answer to this question. In order to evaluate the safety performance and ensure that AVs are safe to be deployed, many driving scenarios need to be created to test the safety performance of prototype AVs using safety metrics and algorithms to measure the metrics in navigated driving scenarios that are being developed as complementary projects. These scenarios are being developed in the open-source CARLA simulation tool that uses Python code and includes scenarios with intersections, vulnerable road users (VRUs), and situations that commonly cause collisions. The creation of these driving scenarios is integral to the overall driving safety performance evaluation methodology development.

Student researcher

Damandeep Singh

Computer science

Hometown: Gurdaspur, Punjab, India

Graduation date: Spring 2023