Development of an Autonomous Driving and Sensors Laboratory Exercise in a senior-level ME/RBE course

Koestner, Bailey

Student Work

Development of an Autonomous Driving and Sensors Laboratory Exercise in a senior-level ME/RBE course

Public Deposited

The primary goals of this project were to test an open-source autonomous driving software for scale cars called Donkey Car and to explore its use as a laboratory in a senior level mechanical and robotics engineering course. The Donkey Car software allows an RC car to be controlled using PWM signals to set throttle and steering angle. Donkey Car collects training images as it is driven storing throttle and steering angle to create a dataset which can be used to generate an autonomous model and can be used for further analysis. A linear regression model is created from the collected data to be uploaded back to the car to enable the autonomous features of the individual cars. The scale cars were fitted with the Arduino microcontroller, a Raspberry PI 4B microprocessor, servo-driver, webcam, portable battery and an inertial measurement unit along with other standard scale car components such as LiPO battery and electronic steering control. The project involved installing the necessary software onto the microprocessor and testing the cars for autonomous driving. This was followed by the development of a detailed installation guide. To further its usability in the classroom each car comes with select sensors to give data read out on acceleration, rotation, and velocity so these values can be used to create an accurate model of the car’s suspension system. A laboratory exercise was created to further outline the goal of the sensor package cars as an educational tool.

This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator