Eric Joseph Wadkins
MIT EECS | MITRE Undergraduate Research and Innovation Scholar
Self-Driving Microscopy: Bayesian Inference for Instrument Tracking and Localization
- Artificial Intelligence and Machine Learning
Dirk R. Englund
The movement of laboratory instruments results in a certain degree of error, which can make precise localization difficult. Techniques such as Bayesian inference allow for an improved localization process — one that takes into account the inherent error arising from the use of physical instruments. Using my previous work for the Quantum Photonics Laboratory, including a real-time system for automated detection and focusing of location-encoded QR codes, I will design and implement a system using Bayesian updating with the aim of improving the existing localization process used by the lab. It is hypothesized that such an improvement will reduce the time required to seek the lab’ s confocal camera.
After completing a UROP project with the Quantum Photonics Laboratory in spring 2017, I wanted to continue my research by working on an extension of my previous work. SuperUROP provided me with the perfect opportunity. I hope to learn new approaches to the problems at hand that will allow my work to have an even greater impact within the lab, all the while making the valuable connections that I’m certain SuperUROP will provide.