UNC CS wins GRAIC autonomous racing competition during CPS‑IoT Week 2021

Abel Karimi and Sridhar Duggirala
Abel Karimi (left) and Parasara Sridhar Duggirala

Doctoral student Abel Karimi and assistant professor Parasara Sridhar Duggirala took first prize in the Generalized Racing Intelligence Competition (GRAIC), a simulated vehicle race affiliated with CPS-IoT Week 2021, the premier event on cyber-physical systems and Internet-of-Things research.

The GRAIC challenges teams of researchers to design a controller for a simulated autonomous vehicle that will enable the vehicle to navigate a track with a variety of obstacles as quickly as possible without colliding with obstacles or going out of bounds. Because the vehicle and track are simulated, the differences in performance are determined by the controlling algorithms.

For GRAIC 2021, due to the COVID-19 pandemic, competitors only had a couple of months to prepare their submissions. The entry from UNC bettered submissions from the University of Illinois at Urbana-Champaign and the University of Pennsylvania. No strangers to autonomous racing success, Karimi and Duggirala were part of a larger UNC CS team that took first place at F1/10, an autonomous race utilizing RC cars that was held during CPS-IoT Week in 2019.

For more information and to watch the GRAIC 2021 live event, visit popgri.github.io/Race.

simulated autonomous lap at GRAIC
Each algorithm was run on the same virtual vehicles and tracks with the results recorded and penalties assessed.
Comments are closed.