F1Tenth Challenge: Autonomous Electric Vehicle
(System Integration Project)

The goal of this project was to take a set of sensors, and actuators, and integrate them to create a remote controllable vehicle that is one-tenth the size of a formula-1 racing car and is capable of piloting itself in limited settings (race track driving). I completed this project with two other students in my Systems Integrations course.

I configured a NVIDIA Jetson Nano microcontroller to interface with multiple sensors and various kinds of motors using Robot Operating System. Part of the configuration process included setting up the microcontroller with a Linux OS and importing the relevant ROS and sensor libraries. We also installed a visualization software that was capable of showing laser scan data (RVIZ) and aiding in SLAM visualization.

 

I used data from a depth camera, a LASER scanner, a VESC 6 MkV Electronic Speed Controller, Inertial Measurement Unit IMU to give feedback to the Spatial Localization and Mapping (SLAM) algorithm that was being used by the microcontroller. By the end of the course we had implemented a number of different autonomous driving algorithms, which included a wall following algorithm, and a center of gap algorithm. The SLAM algorithms that we used were mostly implemented using external libraries and API's