Anomaly Detection during High-Contact Assistive Bathing Task

University of Edinburgh, Statistical Machine Learning and Motor Control Group

January 2024 - Present

Edinburgh, Scotland

Robot Dynamics and Control, Pybullet Simulation Interface, Literature Review, Paper Writing, Classificaiton and Machine Learning, Research Principles, Kuka LBR iiwa Robotic Platform

The goal of this project is to detect anomalies that occur during high-contact human-robot interaction such as an assisted bathing task.

This is a senior design thesis, which means I am responsible for the entirety of the background research, project work, and presentation of the material at the end of the semester.

I implemented the simulated human-robot interface using Pybullet, developed a classification machine learning model to differentiate between safe and unsafe interactions, and performed validation using real-world test data. The project recieved high marks from the thesis committee and will inform future work in the Statistical Learning and Motor Control (SLMC) group at the University of Edinburgh.

This project confirmed for me my passion for research and my desire to further explore robotics, medical applicaitons, and human-robot interaction. I grew in confidence as an independent researcher, learning how to overcome obstacles and think creatively about the problems at hand. I also learned how to be thorough in literature review and thoughtful in the formation of research questions.