Arjun Sharma Poudel’s first-place poster at the 2026 Triangle Student Research Competition highlighted his innovative computational work titled "Machine Learning-Based SEM Image Analysis for Reproducible Characterization of Nanowire Morphology". Co-authored with Carolyn Shi and Dr. Shanthi Iyer at the Joint School of Nanoscience and Nanoengineering, the research addresses a critical bottleneck in semiconductor characterization.

GaAsSb nanowires, grown using molecular beam epitaxy (MBE), hold significant promise for near-infrared photodetection. The ultimate performance of these devices depends heavily on the physical morphology of the nanowires, including their density, diameter, uniformity, merging, and alignment. Traditionally, inspecting these structures via Scanning Electron Microscope (SEM) images has been a subjective, time-consuming process that is difficult to reproduce across large datasets.

To overcome this, the research team developed an automated machine learning workflow designed to extract objective, repeatable measurements directly from SEM image archives.