Oral Presentations
April 15-16, 2026
Interested in CR2C2 activities and engagement opportunities?
Oral Presentations
April 15-16, 2026
CR2C2 | MRI 2 | Project R2-4
April 15, 2026 | 1:55 - 2:15 | Ballroom
Abstract: This project tackled one of the toughest challenges in transportation—making automated and connected vehicles safe and reliable on rural roads. Led by Dr. Subhadeep Chakraborty, the team developed a rural-specific CAV architecture that accounts for poor road markings, limited connectivity, and unpredictable driving conditions. Using advanced AI tools such as the Perception Error Model (PEM), built with a modified FETSGAN network, and GENESIS-RL, a reinforcement learning system for generating realistic edge cases, the research recreated how vehicles perceive and react in complex rural environments. The team validated these models through Vehicle-in-the-Loop tests on instrumented Toyota RAV4 and Lincoln MKZ platforms, linked with digital twin simulations of Tennessee’s rural roadways. Together, these innovations produced measurable safety improvements and a scalable testing framework—paving the way for smarter, safer, and more equitable automated mobility across rural America.
Dr. Subhadeep Chakraborty is an Associate Professor in Mechanical and Aerospace Engineering at the University of Tennessee. Dr. Chakraborty has developed a nationally recognized research program in intelligent transportation systems and connected and automated vehicles (CAVs), with leadership across major projects funded by the U.S. Department of Transportation, NSF, DHS, Oak Ridge National Laboratory, the Tennessee Department of Transportation, and the University Transportation Centers. A central contribution of his research is the design of vehicle-in-the-loop (ViL) simulation platforms that integrate full-scale automated vehicles into high-fidelity digital twin environments, enabling systematic validation of perception, decision-making, and control under rare-event and safety-critical scenarios. His technical expertise spans multi-agent coordination and control, uncertainty quantification for sensor fusion under domain shift, decentralized intersection management, and human factors in transportation.
His research has produced over 40 peer-reviewed journal articles and numerous technical reports, including advances in AI-driven driver impairment detection, adversarial testing of ADS perception stacks, and hybrid physical-digital accelerated testing protocols. Dr. Chakraborty’s leadership extends to coordinating multi-agency collaborations with ORNL, TDOT, city traffic departments, and industry partners, translating research outcomes into policy and operational standards for CAV deployment.
Read more About CR2C2: https://www.cr2c2.com/about-us/about-cr2c2
Read more About FERSC: https://fersc.utk.edu/