Autonomous Electric Go-Kart Racing
Problem statement
Wisconsin Autonomous is using a self-driving go-kart in a racing competition and needs a way to improve and validate the performance of their control algorithms. Autonomous vehicles are already helping to make transportation safer and more efficient [1], but are challenging to deploy in practice due to difficulty in measuring real-world performance in different scenarios [2]. The ability to drive an autonomous vehicle manually while collecting data from on-board sensors will be a huge help to the Wisconsin Autonomous team for both developing their control algorithms and evaluating performance. To address this need, the go-kart will be modified to be human-driven while compliant with evGrand Prix autonomous racing rules [4]. A human driver will safely control the go-kart at handling limits and in novel scenarios so the team can see how sensors measure various aspects of the environment. This data will help verify how closely simulated scenarios match reality to assist sensor and vehicle modeling, and thus control algorithm development [3]. Additionally, the data will provide a way for the autonomous driving performance to be directly and analytically compared to human driving performance. This go-kart will then serve as a reliable test platform for the team and future university research work.
Team members
Alexander Harrison Pletta – leader
Matthew S. Schmidt – communicator
Alexander Paul Bush – accountant
Neo Oh – admin