Randall A. Pietersen

Randall A. Pietersen is a PhD candidate in civil and environmental engineering whose research focuses on creating new deep-learning tools and hyperspectral imaging analysis techniques for USAF airfield damage assessment and the detection of unexploded explosives (UXO). A previous MathWorks Fellowship enabled Randall to validate the principles of his automated sensor calibration methodology and evaluate the process; a second MathWorks Fellowship will support his ongoing efforts to develop synthetic hyperspectral data generation pipelines in support of new spatiospectral machine-learning models that use reflectance-corrected spectral data to detect UXO. He has contributed to numerous MATLAB toolkits, including Deep Learning and Hyperspectral Data Processing. Randall’s research directly addresses the need for improved methods for predictive performance when fielding machine-learning models exposed to limited training data. His work has the potential to advance a broad range of applications of near-surface hyperspectral imaging, from military contexts to mining, agriculture, disaster response, and commercial infrastructure management.