James Dallas is a Ph.D. candidate in the Department of Mechanical Engineering and performs research in the Ersal Research Group. His research focuses on how environmental conditions, such as terrain properties, can be learned online and incorporated into real-time trajectory planning algorithms to enable safe and high-performance autonomous vehicles. To achieve this, he employs nonlinear parameter estimation techniques, such as Unscented Kalman filtering, and neural network modeling to adapt control models in real-time to better capture the characteristics of the vehicle system and environment. He then utilizes these models within a parallel implementation of nonlinear model predictive control on GPUs to improve computational efficiency and make them suitable for real-time implementation. The models and algorithms developed in his research are a critical step in transitioning autonomous vehicles from controlled research settings to the unknown and varying real world in areas such as transportation, space, defense, and agriculture.