As a part of developing unobtrusive wildlife observation, this project aims to deploy an autonomous sea turtle robot to study the behaviors of sea turtles in the wild. My role in this project focuses on developing the software needed to allow the robot to detect and track other turtles. Using custom turtle datasets I compile, I will finetune and compress turtle detection machine learning models that run efficiently on the Raspberry Pi that controls the robot. Further development will include adding machine learning assisted depth perception to determine the robots distance from the turtle it is tracking, to allow the robot to follow and continue observing the targeted turtle.

Through this SuperUROP, I hope to learn a lot about robotics by fully immersing myself in the process of creating a robot by directly working on computer vision software and learning how to apply machine learning to a real world scenario. I also hope to gain more insight into what being a researcher is like, and work on a paper to publish my results.