researchreview UNIVERSITY OF MICHIGAN TRANSPORTATION RESEARCH INSTITUTE
From the Director 2021 was a year of change, adaptation, and growth for UMTRI. Of the more than 12 newly-hired people, five are new assistant-level faculty members and two are new research fellows. dditional growth is on the horizon, as we still have one faculty position and at least one research fellow position to fill. In this edition of the Research Review, we have taken the opportunity to highlight some of the past work of our new faculty and research fellows. In doing so we are highlighting not only their respective achievements, but also making you aware of the exciting new areas in which UMTRI is undertaking research. These areas include artificial intelligence, autonomous vehicle testing, driver behavior, mobility of older adults, pedestrian safety, simulation and modeling, and traffic control systems.
A
We have great expectations for all new members of UMTRI, but we particularly invite you to join us over the course of the next few years in watching the new faculty develop and flourish in their respective areas of research interest. James Sayer UMTRI Director
umtri.umich.edu
APRIL 2022 • VOLUME 4
Intelligent driving intelligence test for autonomous vehicles with naturalistic and adversarial environment FEATURES
Abstract: Driving intelligence tests are critical to the development and deployment of autonomous vehicles. The prevailing approach tests autonomous vehicles in life-like simulations of the naturalistic driving environment. However, due to the high dimensionality of the environment and the rareness of safety-critical events, hundreds of millions of miles would be required to demonstrate the safety performance of autonomous vehicles, which is severely inefficient. We discover that sparse but adversarial adjustments to the naturalistic driving environment, resulting in the naturalistic and adversarial driving environment, can significantly reduce the required test miles without loss of evaluation unbiasedness. By training the background vehicles to learn when to execute what adversarial maneuver, the proposed environment becomes an intelligent environment for driving intelligence testing. We demonstrate the effectiveness of the proposed environment in a highwaydriving simulation. Comparing with the naturalistic driving environment, the proposed environment can accelerate the evaluation process by multiple orders of magnitude.
Citation: Feng, S., Yan, X., Sun, H., Feng, Y., & Liu, H. X. (2021). Intelligent driving intelligence test for autonomous vehicles with naturalistic and adversarial environment. Nature communications, 12(1), 1-14. https://www.nature.com/articles/s41467021-21007-8 Continued on page 2