GCSP research stipend | Spring 2025

Optimizing Turning Behavior and Lighting Conditions to Reduce Carbon Emissions in Urban Driving

In this project, Duckietown, an ecosystem for AI driving and research, was successfully installed. Additionally, a Duckiebot was built, and the Duckiematrix simulator was set up. Using the simulator, real-world driving data was collected to simulate urban environments. This data included parameters such as fuel consumption, speed, acceleration, stop time, and braking, with a focus on conditions like low lighting and turning behavior. By analyzing this data, the study evaluates how different driving behaviors, particularly in low-visibility environments and during turns, impact fuel consumption. The findings aim to enhance energy-efficient driving strategies and minimize overall energy loss.