Kevin Potter
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kevin Potter has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
Albuquerque Sandia National Laboratories (Snl-Nm)
5 shared publicationsUsdoe National Nuclear Security Administration (Nnsa)
3 shared publications- Dan Bolintineanu
Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
2 shared publications - David Erik Kittell
Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
2 shared publications - David P. Adams
Sandia National Laboratories 1 , Albuquerque, New Mexico 87185
2 shared publications - Michael Joseph Kim Abere
Sandia National Laboratories , Albuquerque, New Mexico 87123
2 shared publications
Latest Publications
- 1 October 2026
- Published by Office of Scientific and Technical Information (OSTI)
SAM-I-AM: Segment Anything Model in Additive Manufacturing
Optical imaging in advanced manufacturing offers a promising path toward robust on-machine inspection, defect detection, and metrology, with the potential to increase production efficiency and improve quality control. However, image analysis often remains manual because robust segmentation—the assignment of image pixels to objects or features of interest—is challenging across different builds, machines, camera configurations, and part geometries. This work evaluates automated segmentation of lay
- 1 October 2026
- Published by Office of Scientific and Technical Information (OSTI)
SAM-I-AM: Segment Anything Model in Additive Manufacturing
Optical imaging in advanced manufacturing offers a promising path toward robust on-machine inspection, defect detection, and metrology, with the potential to increase production efficiency and improve quality control. However, image analysis often remains manual because robust segmentation—the assignment of image pixels to objects or features of interest—is challenging across different builds, machines, camera configurations, and part geometries. This work evaluates automated segmentation of lay
- Open Access
- 22 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19 (19391404) , 15071-15089
VoroClust: Scalable Clustering for Remote Sensing
Although supervised machine learning provides a powerful framework for image classification and segmentation, it requires comprehensive, consistent data sets, which are not available for many remote-sensing applications. Remote-sensing datasets are expensive to collect, and each is acquired under different environmental conditions, or with significant variations in system operating parameters. Unsupervised clustering algorithms analyze the structure of each dataset independently, rather than drawing on similarities with existing ‘training’ examples, and are thus well suited for practical remote-sensing applications. We introduce VoroClust, a fast, density-based unsupervised clustering algorithm applicable to high-resolution and highdimensional data. VoroClustruns as fast as distance-based clustering methods, while capturing complex regional geometries at least as well as current density-based methods. It uses a datacentered sphere cover to reduce computational demands, while still capturing data topology. It then propagates clusters outward from local peaks in density. We show that VoroClustprovides fast, state-of-the-art clustering for both high-resolution polarimetric synthetic aperture radar (PolSAR) and high-dimensional hyperspectral imaging (HSI) datasets.
- 22 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19 (19391404) , 15071-15089
Open AccessVoroClust: Scalable Clustering for Remote Sensing
Although supervised machine learning provides a powerful framework for image classification and segmentation, it requires comprehensive, consistent data sets, which are not available for many remote-sensing applications. Remote-sensing datasets are expensive to collect, and each is acquired under different environmental conditions, or with significant variations in system operating parameters. Unsupervised clustering algorithms analyze the structure of each dataset independently, rather than drawing on similarities with existing ‘training’ examples, and are thus well suited for practical remote-sensing applications. We introduce VoroClust, a fast, density-based unsupervised clustering algorithm applicable to high-resolution and highdimensional data. VoroClustruns as fast as distance-based clustering methods, while capturing complex regional geometries at least as well as current density-based methods. It uses a datacentered sphere cover to reduce computational demands, while still capturing data topology. It then propagates clusters outward from local peaks in density. We show that VoroClustprovides fast, state-of-the-art clustering for both high-resolution polarimetric synthetic aperture radar (PolSAR) and high-dimensional hyperspectral imaging (HSI) datasets.
- Open Access
- 9 June 2025
- Published by AIP Publishing in Journal of Applied Physics
- Vol. 137 (22)
Continuum shock mixture models for Ni+Al multilayers: Inert mesoscale simulations
Mesoscale modeling of shock waves in Ni+Al multilayers poses significant challenges that are due, in part, to shock-induced chemical reactions. Current modeling approaches utilize reactive molecular dynamics (MD), but they are limited to resolving domains of only a few hundred nanometers. In contrast, actual multilayer superlattices can be tens of micrometers thick, and they exhibit non-ideal (i.e., wavy) interfaces. The second part of our research builds upon previous work developing physically based, thermodynamically complete equations of state for various Ni and Al intermetallic compositions. Here, we introduce a novel workflow for high-fidelity mesoscale simulations of Ni+Al multilayers using a continuum hydrocode. By increasing the simulation domain size beyond MD limitations (e.g., 2×6μm2) and incorporating explicit interfacial roughness, we investigate the shock response of Ni+Al multilayers at previously unexplored scales. Our experimental design encompasses nine multilayer geometries with varying roughness amplitudes and tilt angles (θ=15°, 30°, and 45°), alongside 19 flyer impact velocities ranging from 0.3 to 3.0 km/s, resulting in a total of 171 high-fidelity simulations. The bulk shock state from inert 2D mesoscale simulations aligns with the law of mixtures, while temperature and pressure fluctuations strongly correlate with multilayer geometry types. A new metric dubbed the “hot spot probability integral” shows a greater dependence on a tilt angle than interfacial roughness.
- 9 June 2025
- Published by AIP Publishing in Journal of Applied Physics
- Vol. 137 (22)
Open AccessContinuum shock mixture models for Ni+Al multilayers: Inert mesoscale simulations
Mesoscale modeling of shock waves in Ni+Al multilayers poses significant challenges that are due, in part, to shock-induced chemical reactions. Current modeling approaches utilize reactive molecular dynamics (MD), but they are limited to resolving domains of only a few hundred nanometers. In contrast, actual multilayer superlattices can be tens of micrometers thick, and they exhibit non-ideal (i.e., wavy) interfaces. The second part of our research builds upon previous work developing physically based, thermodynamically complete equations of state for various Ni and Al intermetallic compositions. Here, we introduce a novel workflow for high-fidelity mesoscale simulations of Ni+Al multilayers using a continuum hydrocode. By increasing the simulation domain size beyond MD limitations (e.g., 2×6μm2) and incorporating explicit interfacial roughness, we investigate the shock response of Ni+Al multilayers at previously unexplored scales. Our experimental design encompasses nine multilayer geometries with varying roughness amplitudes and tilt angles (θ=15°, 30°, and 45°), alongside 19 flyer impact velocities ranging from 0.3 to 3.0 km/s, resulting in a total of 171 high-fidelity simulations. The bulk shock state from inert 2D mesoscale simulations aligns with the law of mixtures, while temperature and pressure fluctuations strongly correlate with multilayer geometry types. A new metric dubbed the “hot spot probability integral” shows a greater dependence on a tilt angle than interfacial roughness.
- Open Access
- 1 September 2024
- Published by Office of Scientific and Technical Information (OSTI)
Scoring Morphological Change in Computed Tomography Images of Pentaerythritol Tetranitrate (PETN)
- 1 September 2024
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessScoring Morphological Change in Computed Tomography Images of Pentaerythritol Tetranitrate (PETN)
- Open Access
- 1 June 2024
- Published by Office of Scientific and Technical Information (OSTI)
Graph Convolutional Neural Networks as Surrogate Models for Climate Simulation
- 1 June 2024
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessGraph Convolutional Neural Networks as Surrogate Models for Climate Simulation