Date Thesis Awarded

5-2022

Access Type

Honors Thesis -- Access Restricted On-Campus Only

Degree Name

Bachelors of Science (BS)

Department

Physics

Advisor

Irina Novikova

Committee Members

David Armstrong

Stephen Trefethen

Abstract

We combine a camera, an optical fiber, and artificial intelligence into a single optical magnetometer for magnetic field measurement. This novel combination provides enhanced spatial resolution, a mobile configuration, and efficient, unbiased data processing capabilities. The magnetometer is based on an undergraduate laboratory Faraday rotation apparatus (a glass rod surrounded by a solenoid), linked to a camera via a multimode optical fiber. To identify varying magnetic field strengths, an image classification algorithm analyzes the fiber output “speckle” patterns that result from different magnetically-induced changes in probe beam polarization. Initially, as we constructed and strengthened the algorithm, we simulated these polarization changes using a waveplate, and we investigated the algorithm’s response to external factors such as natural fluctuations in probe beam, and therefore image, intensity. Later, we replaced the waveplate with the glass rod and solenoid. Ultimately, we created a sensor with angular sensitivity as small as 0.6 x 10−4 degrees, corresponding to magnetic fields on the order of 0.5 μT. The device is applicable in structural defects detection, including settings that require small size and mobility and that might not
be electronics-friendly.

Recommended Citation

Brown, Sofia, "Optical Fiber-Linked Magnetometer Employing Artificial Intelligence for Magnetic Field Measurement" (2022). Undergraduate Honors Theses. William & Mary. Paper 1801.
https://scholarworks.wm.edu/honorstheses/1801