The Mushr project involved transforming a standard remote-controlled car by equipping it with a specialized board and necessary sensors to enable autonomous driving. Our team was tasked with developing the software to autonomously pilot the vehicle. This intricate process involved several key phases: firstly, implementing localization techniques to accurately determine the vehicle's position; secondly, developing control algorithms for precise movement management; thirdly, creating navigation plans for obstacle avoidance; and fourthly, conducting extensive tests in a simulated environment to ensure functionality and reliability. The culmination of our efforts was the real-world application, where the vehicle successfully navigated through a building with various obstacles, demonstrating the effectiveness of our software solutions.
Map of the room with the cars approximate location. The goal was to get as close to the center of the L shape as possible.
PID controller running in the simulator used to control the variables used in the PID control algorithm. This picture was captured after PID controller algorithm was tuned
A look at the realtime planning of the Mushr car through the simulator program.
In the Mushr autonomous vehicle project, I served as the lead localization developer, responsible for spearheading the development of precise localization techniques that enabled the car to accurately determine its position within its environment. I contributed as a general developer across all other phases of the project, including control algorithms, planning strategies, simulation testing, and the final implementation of driving the vehicle autonomously through real-world obstacles. My role was integral to ensuring the project's success across its multiple, complex stages.