Vehicle Autonomy & the EcoCAR Challenge

Cameras see texture; radar sees through weather and darkness. AICIP’s autonomous-vehicle research made that complementarity concrete: CenterFusion (WACV 2021) introduced a middle-fusion approach that associates radar detections with camera-based object centers for 3D detection and velocity estimation — work that became a widely used baseline in radar–camera fusion — alongside radar region-proposal networks (RRPN) and multi-object tracking work.

The applied home for this research is the U.S. Department of Energy’s EcoCAR advanced-vehicle technology competitions, administered by Argonne National Laboratory. Dr. Qi served as a faculty PI for UT’s team in the EcoCAR Mobility Challenge (2018–2022), where students integrated connected-and-automated-vehicle features into a production vehicle. In 2026, UT was selected as one of twenty North American universities for the EcoCAR Innovation Challenge (fall 2026–spring 2030), competing on the Stellantis track with a hybrid 2026 Jeep Cherokee; Dr. Qi is one of three faculty advisors, with AICIP contributing the perception and human–machine interaction expertise.