Implementation of Machine Learning and ROS into Self-Driving RC Cars

Anderson Molter, Sanjay Sarma, and Dr. Ramviyas Parasuraman, Department of Computer Science, University of Georgia, 803 Boyd Graduate Studies Research Center, Athens GA, 30602

We present a project which involves creating a simulation and hardware framework for an innovative minimalistic self-driving car with AI capabilities for aiding with teaching Robotics and AI curriculum at Universities. In this work, we have designed guidelines for converting a simple low-cost RC car into an AI-enabled device. The car uses an array of 5 IR sensors to detect the tracks. Along with this hardware enhancement, a simulation software in Unity is created of this RC car using the same algorithm to allow students to dive deep into AI algorithms and see its many applications and means of operation. Our current development efforts include the integration of Software and Hardware layers. Also, we will be using libraries like OpenCV to process the camera image data and design a machine learning model around it to allow the car to learn from its mistakes, successes, and many other types of interactions with the environment. In this presentation, we show the hardware design and development guidelines, Unity software for simulation testing in a game-like setup, and their integration for AI Education.

Additional Abstract Information

Presenter: Anderson Molter

Institution: University of Georgia

Type: Poster

Subject: Computer Science

Status: Approved

Time and Location

Session: Poster 5
Date/Time: Tue 12:30pm-1:30pm
Session Number: 4050