Control System Integration for Automating an Electric Golf Cart

Millions of people lose their lives each year, due to traffic accidents. Autonomous vehicles could not only help prevent these accidents, but create a technological renaissance. It's well known that autonomous cars are expected to be safer, more efficient, and the future of campus mobility. College students as a future workforce need to be prepared in these autonomous vehicle fields. Electric golf carts, with much simpler systems than a real car, are a perfect choice for education. The purpose of this project is to convert an electric Club Car golf cart into a test platform for deep learning algorithms. The deep learning algorithms for path planning/searching and self-driving will eventually lead to an autonomous golf cart that can navigate the campus of Eastern Michigan University. We will integrate peripheral sensors such as Lidar, GPS, IMUs, and control systems to the golf cart. The original golf cart's mechanical chassis is the control system's foundation. Sensors and control systems used to automate this golf cart will be built as an extension of the existing systems on the chassis. This paper will focus on the initial design considerations and process for converting the original golf cart into a remote-controlled transportation system. This work will be used for testing digital control of the golf cart's steering, brake and acceleration systems. The paper will explore challenges and design constraints the team encounters during this process. In addition, environmental data collected from peripheral sensors, path planning and searching, lane detection, and power management will also be discussed.