This report explores the feasibility of creating an autonomous 1:10 scale car capable of racing against RC cars on any indoor track. Custom made and equipped with a livestream sensor display module, the car leverages Artificial Neural Networks (ANN) to produce optimal driving outputs. Manipulating grayscale pixel input from a single, front-facing camera, the car navigates without the use of mapping or localization. Instead, the car relies on extensive training that allows the vehicle to adapt to variable conditions.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.
Electrical and Computer Engineering