AI in Robotics: Implementing Donkey Car

Document Type


Publication Date



Computer Sciences


Robotic System is a diverse infrastructure in the modern world whose application is prominent in the automobile industry. One such application includes the self-driving cars that are slowly being available to the general public here in the US. The development of autonomous driving vehicles has been exponentially increasing due to the rapid integration of artificial technology with robotics. To demonstrate this similar mechanism, this research conducts a demo using Donkey Car - a small-sized supervised learning Radio Controlled (RC) car consisting of Raspberry Pi, which is used to replicate the driving functionality of the autonomous vehicles on a closed track in autopilot. The donkey car is trained by raw images and positioning metadata of the path using the TensorFlow and OpenCV python module. The main objective for the device will be to remain in between the closed-track bounded by lines while maintaining and making its own driving decision and driving through turns and cruise, mimicking human behavior.

First Advisor

Santosh KC

Research Area

Computer Science

This document is currently not available here.