The UCLA RoMeLa lab is a leading robotics research lab with an emphasis on humanoid and novel robotic locomotion. My initial work at the lab primarily focused on helping with development and assembly of small scale autonomous cars. The goal of this project was to use an existing vehicle platform and modify it to be robo-centric in such a way that it could beat a world champion. This was achieved by using a fusion of behavioral cloning to train the model and deep reinforcement learning ran in real time to optimize the vehicle’s trajectory.

I helped design & test a raised camera mount, as the original camera mount put the camera too low to obtain accurate positioning data. This was a challenge as all the camera and cooling had to be packaged onto the same part and the height was constrained by the body of the vehicle.

After the student leading research for this project graduated I moved over to help with other projects including optimization for the BALLU buoyant robot’s end effectors and designing and building the structure for an environment for testing of the LIMMS multi-agent modular delivery robot system.