Researchers at Waterloo Engineering and in Japan have developed a mathematical method to improve the safety of autonomous vehicles and other autonomous systems.
The new tool, developed as part of a five-year collaboration, gives engineers a systematic way to determine the safety margin made necessary by uncertainties in sensors and artificial intelligence (AI) algorithms.
In the case of autonomous vehicles, for instance, extra braking distance may be required to guarantee safety because systems can perceive objects to be further away than they actually are.
New method submitted for patent
Krzysztof Czarnecki, a professor of electrical and computer engineering who is director of the Intelligent Systems Engineering Lab, led the Waterloo team.
It is backed by $1.2 million in funding over five years from an agency of the Japanese government. Waterloo is the only foreign institution working on the project with researchers at the National Institute of Informatics in Tokyo.
The new mathematical method, which has been submitted for a Japanese patent, was to be presented today at the virtual 13th NASA Formal Methods Symposium.