Mcity develops safety testing methodology, scoring approach for highly automated vehicles
Mcity at the University of Michigan has developed a testing methodology for highly automated vehicles, or HAVs, within a closed test facility to ensure their safety before public road testing and deployment. Such tests could be used to create an independent assessment of the safety performance of HAVs, which is a crucial element for improving the public’s trust in automated vehicle technology.
Mcity outlines the primary challenges of conducting these tests, a methodology for completing them, and a concept for using the results to develop a safety rating system in a white paper published today. Mcity is a public-private partnership led by U-M to accelerate advanced mobility vehicles and technologies.
HAVs, for this paper, refers to Level 4 and Level 5 automation, as defined by SAE International. The testing methodology was developed to expand on the Mcity ABC Test concept Mcity introduced last year. The ABCs of the three-part test include Accelerated evaluation, Behavior competence and Corner cases. The new white paper focuses on the behavior competence tests, but the methodology in large part also applies to accelerated evaluation and corner case tests.
Accelerated evaluation concentrates on the most common risky driving situations. Behavior competence puts vehicles through a set of scenarios that correspond to major motor vehicle crash frequency. Corner-case testing focuses on situations that test the limits of automated vehicle performance and technology.
The testing methodology and concept for scoring HAV performance comes as consumer confidence in automated vehicles continues to decline. The J.D. Power Mobility Confidence Index for self-driving vehicles, released earlier this month, decreased for the first time —to 35 from 36 on a 100-point scale—for American consumers and to 36 from 39 for Canadian consumers. The study, conducted with Survey Monkey, surveyed more than 8,500 consumers in March 2020.
The white paper is titled, “Conducting the Mcity ABC Test: A Testing Method for Highly Automated Vehicles.”