Overview
Kyle Bradbury is the Managing Director of the Energy Data Analytics Lab at the Duke University Energy Initiative. He brings experience in machine learning and statistical modeling to energy problems. He completed his Ph.D. at Duke University, with research focused on modeling the reliability and cost trade-offs of energy storage systems for integrating wind and solar power into the grid. Kyle holds a M.S. in Electrical Engineering from Duke University where he specialized in statistical signal processing and machine learning, and a B.S. in Electrical Engineering from Tufts University. He has worked for ISO New England, MIT Lincoln Laboratories, and Dominion.
Current Appointments & Affiliations
Assistant Research Professor in the Department of Electrical and Computer Engineering
·
2020 - Present
Pierre R. Lamond Department of Electrical and Computer Engineering,
Pratt School of Engineering
Assistant Research Professor in the Division of Environmental Social Systems
·
2024 - Present
Environmental Social Systems,
Nicholas School of the Environment
Faculty Fellow in the Nicholas Institute for Energy, Environment & Sustainability
·
2022 - Present
Nicholas Institute for Energy, Environment & Sustainability,
University Institutes and Centers
Recent Publications
Machine learning inversion of interatomic force constants from single-crystal inelastic neutron scattering
Journal Article Digital Discovery · April 1, 2026 Atomic vibrations govern many macroscopic properties of materials, but experiments to comprehensively probe them remain challenging. Inelastic neutron scattering (INS) is a powerful technique to map phonon dispersions in crystals, especially when leveragin ... Full text CiteSegment anything, from space?
Conference Proceedings 2024 IEEE Winter Conference on Applications of Computer Vision Wacv 2024 · January 3, 2024 Recently, the first foundation model developed specifically for image segmentation tasks was developed, termed the "Segment Anything Model"(SAM). SAM can segment objects in input imagery based on cheap input prompts, such as one (or more) points, a boundin ... Full text CiteRandomized Histogram Matching: A Simple Augmentation for Unsupervised Domain Adaptation in Overhead Imagery
Journal Article IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · January 1, 2024 Modern deep neural networks (DNNs) are highly accurate on many recognition tasks for overhead (e.g., satellite) imagery. However, visual domain shifts (e.g., statistical changes due to geography, sensor, or atmospheric conditions) remain a challenge, causi ... Full text CiteRecent Grants
IUCRC Phase I Duke University: Center for Innovation in Risk-analysis for Climate Adaption and Decision-making (CIRCAD)
ResearchParticipating Faculty Member · Awarded by National Science Foundation · 2025 - 2030Climate TRACE Phase 8
ResearchPrincipal Investigator · Awarded by WattTime · 2026 - 2027Climate TRACE Phase 7
ResearchPrincipal Investigator · Awarded by WattTime · 2025 - 2026View All Grants
Education
Duke University ·
2013
Ph.D.