Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Vision”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Cooling Matters: Benchmarking Large Language Models and Vision-Language Models on Liquid-Cooled Versus Air-Cooled H100 GPU Systems

The unprecedented growth in artificial intelligence (AI) workloads, recently dominated by large language models (LLMs) and vision-language models (VLMs), has intensified power and cooling demands in data centers. This study benchmarks LLMs and VLMs on two HGX nodes, each with 8× NVIDIA H100 graphics processing units (GPUs), using liquid and air cooling. Leveraging GPU Burn, Weights & Biases, and IPMItool, we collect detailed thermal, power, and computation data. Results show that the liquid-cooled systems maintain GPU temperatures between 41-50$^\circ$C, while the air-cooled counterparts fluctuate between 54-72$^\circ$C under load. This thermal stability of liquid-cooled systems yields 17% higher performance (54 TFLOPs/ GPU vs. 46 TFLOPs/GPU), performance-per-watt, reduced energy overhead, and greater system efficiency than the air-cooled counterparts. These findings underscore the energy and sustainability benefits of liquid cooling, offering a compelling path forward for hyperscale data centers seeking to optimize AI infrastructure. https://github.com/iscaas/Cooling-Matters.

Latif, Imran↗
DOE OSTI2023

UVCGAN: UNet Vision Transformer cycle-consistent GAN for unpaired image-to-image translation

Unpaired image-to-image translation has broad applications in art, design, and scientific simulations. One early breakthrough was CycleGAN that emphasizes one-to-one mappings between two unpaired image domains via generative-adversarial networks (GAN) coupled with the cycle-consistency constraint, while more recent works promote one-to-many mapping to boost diversity of the translated images. Motivated by scientific simulation and one-to-one needs, this work revisits the classic CycleGAN framework and boosts its performance to outperform more contemporary models without relaxing the cycle-consistency constraint. To achieve this, we equip the generator with a Vision Transformer (ViT) and employ necessary training and regularization techniques. Compared to previous best-performing models, our model performs better and retains a strong correlation between the original and translated image. An accompanying ablation study shows that both the gradient penalty and self-supervised pre-training are crucial to the improvement. To promote reproducibility and open science, the source code, hyperparameter configurations, and pre-trained model are available at https: //github.com/LS4GAN/uvcgan.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

VDiSC: An Open Source Framework for Distributed Smart City Vision and Biometric Surveillance Networks

Recent global growth in the interest of smart cities has led to trillions of dollars of investment toward research and development. These connected cities have the potential to create a symbiosis of technology and society and revolutionize the cost of living, safety, ecological sustainability, and quality of life of societies on a world-wide scale. Some key components of the smart city construct are connected smart grids, self-driving cars, federated learning systems, smart utilities, large-scale public transit, and proactive surveillance systems. While exciting in prospect, these technologies and their subsequent integration cannot be attempted without addressing the potential societal impacts of such a high degree of automation and data sharing. Additionally, the feasibility of coordinating so many disparate tasks will require a fast, extensible, unifying framework. To that end, we propose the Distributed Smart City framework for Vision, or VDiSC. VDiSC serves as a unified biometric API harness that allows for seamless evaluation, deployment, and simple pipeline creation for heterogeneous biometric software. VDiSC additionally provides a fully declarative capability for defining and coordinating custom machine learning and sensor pipelines, allowing the distribution of processes across otherwise incompatible hardware and networks. VDiSC ultimately provides a way to quickly configure, hot-swap, and expand large coordinated or federated systems online without interruptions for maintenance. Because much of the data collected in a smart city contains Personally Identifying Information (PII), VDiSC also provides built-in tools and layers to ensure secure and encrypted streaming, storage, and access of PII data across distributed systems.

Brogan, Joel↗
DOE OSTI2024

In Silico Human Mobility Data Science: Leveraging Massive Simulated Mobility Data (Vision Paper)

Human mobility data science using trajectories or check-ins of individuals has many applications. Recently, we have seen a plethora of research efforts that tackle these applications. However, research progress in this field is limited by a lack of large and representative datasets. The largest and most commonly used dataset of individual human trajectories captures fewer than 200 individuals, while datasets of individual human check-ins capture fewer than 100 check-ins per city per day. Thus, it is not clear if findings from the human mobility data science community would generalize to large populations. Since obtaining massive, representative, and individual-level human mobility data is hard to come by due to privacy considerations, the vision of this work is to embrace the use of data generated by large-scale socially realistic microsimulations. Informed by both real data and leveraging social and behavioral theories, massive spatially explicit microsimulations may allow us to simulate entire megacities at the person level. The simulated worlds, which do not capture any identifiable personal information, allow us to perform “in silico” experiments using the simulated world as a sandbox in which we have perfect information and perfect control without jeopardizing the privacy of any actual individual. In silico experiments have become commonplace in other scientific domains such as chemistry and biology, permitting experiments that foster the understanding of concepts without any harm to individuals. This work describes challenges and opportunities for leveraging massive and realistic simulated alternate worlds for in silico human mobility data science.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

ORBIT-2: Scaling Exascale Vision Foundation Models for Weather and Climate Downscaling

Sparse observations and coarse-resolution climate models limit effective regional decision-making, underscoring the need for robust downscaling. However, existing AI methods struggle with generalization across variables and geographies and are constrained by the quadratic complexity of Vision Transformer (ViT) self-attention. We introduce ORBIT-2, a scalable foundation model for global, hyper-resolution climate downscaling. ORBIT-2 incorporates two key innovations: (1) Residual Slim ViT (Reslim), a lightweight architecture with residual learning and Bayesian regularization for efficient, robust prediction; and (2) TILES, a tile-wise sequence scaling algorithm that reduces self-attention complexity from quadratic to linear, enabling long-sequence processing and massive parallelism. ORBIT-2 scales to 10 billion parameters across 65,536 GPUs, achieving up to 4.1 ExaFLOPS sustained throughput and 74–98% strong scaling efficiency. It supports downscaling to 0.9 km global resolution and processes sequences up to 4.2 billion tokens. On 7 km resolution benchmarks, ORBIT-2 achieves high accuracy with R2 scores in range of 0.98–0.99 against observation data.

Wang, Xiao [ORNL] (ORCID:0000000165451943)↗
DOE OSTI2024

A Vision for the Science of Rare Isotopes

The field of nuclear science has considerably advanced since its beginning just over a century ago. Today, the science of rare isotopes is on the cusp of a new era with theoretical and computing advances complementing experimental capabilities at new facilities internationally. In this article we present a vision for the science of rare isotope beams (RIBs). We do not attempt to cover the full breadth of the field; rather, we provide a perspective and address a selection of topics that reflect our own interests and expertise. We focus in particular on systems near the drip lines, where one often finds nuclei that are referred to as exotic and where the role of the nuclear continuum is only just starting to be explored. An important aspect of this article is its attempt to highlight the crucial connections between nuclear structure and the nuclear reactions required to fully interpret and leverage the rich data to be collected in the next years at RIB facilities. Further, we connect the efforts in structure and reactions to key questions of nuclear astrophysics.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2024

Hyperparameter Studies for Vision Transformers Trained on High-Fidelity Simulations

This library is a collection of python modules that define, train, and analyze vision-transformer (ViT) machine learning models. The code implements, with mild modifications, ViT models that have been made publicly available through publication and GitHub code. The training data for these models is hydrodynamic simulation output in the form of numpy arrays. This library contains code to train these ViT models on the hydrodynamic simulation output with a variety of hyperparameters, and to compare the results of such models. Furthermore, the library contains definitions of simple convolutional neural network (CNN) machine learning architectures which can be trained on the same hydrodynamic simulation output. These are included as a reference point to compare the ViT models to. Additionally, the library includes trained ViT and CNN models and example input data for demonstration purposes. The code is based on the PyTorch python library.

Callis, Skylar↗
DOE OSTI2024

Effective Visualization of Radar Data for Users Impacted by Color Vision Deficiency

Color vision deficiency (CVD) is a decreased ability to discern between particular colors. Eight percent of genetic males and half a percent of genetic females have some form of CVD, with many in the radar community falling into this group. When presenting data on a two-dimensional plane, it is common to use colors to represent values via a colormap. Colormap choice in the radar community is influenced by the ability to highlight scientifically interesting features in data, institutional choices, and domain dominance of legacy colormaps. The problem with these current colormaps is that many do not project well for those with CVD (i.e., green next to red). In working with the CVD community to address this problem, multiple colormaps for moments such as equivalent reflectivity factor and Doppler velocity were created for users with forms of CVD such as deuteranomaly, protanomaly, protanopia, and deuteranopia using Python tools such as colorspacious and viscm. We show how these colormaps can improve interpretability for four cases: a mesoscale convective system, a pyrocumulonimbus storm, a wintertime midlatitude cyclone, and widespread storms with a large bird migration. These new radar equivalent reflectivity factor, Doppler velocity, and polarization colormaps are designed to highlight rain, frozen precipitation, nonmeteorological targets, and velocity-based items, are perceptually uniform, and are visually friendly for those with CVD.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2014

Human microbiome science: vision for the future, Bethesda, MD, July 24 to 26, 2013

A conference entitled ‘Human microbiome science: Vision for the future’ was organized in Bethesda, MD from July 24 to 26, 2013. The event brought together experts in the field of human microbiome research and aimed at providing a comprehensive overview of the state of microbiome research, but more importantly to identify and discuss gaps, challenges and opportunities in this nascent field. This report summarizes the presentations but also describes what is needed for human microbiome research to move forward and deliver medical translational applications.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Advancing subsurface analysis: Integrating computer vision and deep learning for the near real-time interpretation of borehole image logs in the Illinois Basin-Decatur Project

The accurate quantification and mapping of subsurface natural fracture systems using borehole imaging logs are critical for the success of CO 2 sequestration in geologic formations, optimization of engineered geothermal systems, and hydrocarbon production enhancement. However, traditional interpretation processes suffer from time-consuming procedures and human bias. To address these challenges and expedite fracture analysis, we investigated the application of integrated computer vision and DL workflows to automate image log analysis. Specifically, the design of our workflow was crafted to swiftly detect fractures and baffles by using actual electrical resistivity of borehole wall from microresistivity imaging device alongside their binary representation. This novel approach significantly reduces computational time while providing invaluable insights. By incorporating conventional logging and microseismic data, we present a regional subsurface natural fracture mapping technique. Through the minimization of human bias in image log analysis, our automated workflow achieves reduced fracture interpretation time and costs while ensuring robust and reproducible results. We demonstrated the efficacy of our approach by applying the workflow to the Illinois Basin-Decatur Project site. The automated workflow successfully identified major fractured zones, multiple baffles, and an interbedded layer with a high resolution of 0.01 ft or 0.12 in. (0.3 cm) and can be upscaled to any desired resolution. Validation through microseismic and image log interpretations allows for accurate and near-real-time mapping of fractures and baffles, significantly enhancing CO 2 pressure forecasting and postinjection site care. Our approach stands out due to its robustness, consistency, and reduced computational cost compared with alternative feature extraction technologies. It presents exciting possibilities for advancing CO 2 sequestration and engineered geothermal efforts by offering comprehensive and efficient fracture mapping solutions. This technology can contribute significantly to the optimization of CO 2 sequestration projects, facilitating sustainable environmental practices, and combating climate change.

Geochemistry & Geophysics↗
DOE OSTI2024

End State: Vision for Future Land Use at Hanford

An end state vision is influenced by practical constraints associated with technology limitations, resource availability, costs, timing, and unintended consequences of remedial actions, and a site’s next intended use. Therefore, subsurface environmental remediation goals often need to be balanced against the risks and costs of contaminant removal versus leaving contamination in place. For example, in regions where contaminant concentrations are high but immobile, the intended site use should be evaluated in conjunction with the risk of re-mobilization with an aggressive remedy. By contrast, the technological challenges, cost, and time to reach remedial objectives of cleaning up low levels of contamination distributed over large areas in the subsurface may require different considerations for a site’s next intended use.

environmental remediation, environmental managemen↗
DOE OSTI2024

TEAMER: Performance mapping of Re Vision's Persistance PTO

Contains datasets from experimental measurements that were used to validate Re Vision's Persistence PTO's efficiency and performance. These measurements were obtained using a dynamometer test bench setup. The data includes open-circuit voltage and loss measurements to validate machine characteristics, efficiency mapping tests to determine the generator's performance mapping, and efficiency mapping tests to determine the converter's efficiency over the feasible operating range. This data was collected between June 2023 and September 2023. The data was collected at the National Renewable Energy Laboratory's Flatirons Campus, Colorado, United States. The data was collected using NREL's 5-kW dynamometer test bench, equipped with a torque sensor and various voltage and current sensors fed to a dedicated data acquisition system. Units for the data are included in the data file headers for each data series. A text editor or spreadsheet software such as Excel is required to view the *.csv data. The data are also provided in *.mat files. To view data plots, a Matlab script with *.mat files are provided.

16 TIDAL AND WAVE POWER↗
DOE OSTI2021

Computer Vision Techniques Used to Monitor the Alignment of Cavities and Solenoids in the PIP-II Prototype SSR1 Cryomodule

The alignment of the SRF PIP-II string components is studied as the acceptable beam deflection, offset and defocusing, which may otherwise cause beam loss. Simulations and measurements established that the maximum deviation of the beam pipe from the reference orbit should not exceed a small fraction of the beam aperture. To observe the translations and rotations of each single component within the cryomodule, optical instruments (H-BCAM) surveying highly reflective targets, installed in the internal assembly of the module were used. The alignment monitoring concept for the PIP II SSR1 prototype cryomodule, along with relevant measurements of the components’ position monitoring during coldmass cooldown is presented in this contribution. This development paves the way to new computer vision applications in the field of cryomodule assemblies in cleanroom environment, in which robotically-assisted operations have the potential to dramatically reduce the risk of chemical and particulate contamination.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Integration of computer vision system to track the alignment SRF cavities into the test cryostat for PIP-II at Fermilab

PIP-II cryomodules use a computer vision system (H-BCAMs system) to monitor the alignment of SRF cavities and focusing lenses during assembly, testing, and operation. This contribution details the integration of the H-BCAMs into the Spoke Test Cryostat (STC) at Fermilab, which is utilized for cold testing SRF cavities prior to their integration into the string assembly. Thermal and structural finite element analyses were employed to estimate the cavities’ deformations, to be validated during cold testing in the STC using H-BCAMs. Notably, this marks the first instance of H-BCAMs integration into a cryostat and operation within a cryogenic environment.

43 PARTICLE ACCELERATORS↗
DOE OSTI2013

Sensitivity of Utility-Scale Solar Deployment Projections in the SunShot Vision Study to Market and Performance Assumptions

The SunShot Vision Study explored the potential growth of solar markets if solar prices decreased by about 75% from 2010 to 2020. The ReEDS model was used to simulate utility PV and CSP deployment for this present study, based on several market and performance assumptions - electricity demand, natural gas prices, coal retirements, cost and performance of non-solar renewable technologies, PVresource variability, distributed PV deployment, and solar market supply growth - in addition to the SunShot solar price projections. This study finds that utility-scale solar deployment is highly sensitive to solar prices. Other factors can have significant impacts, particularly electricity demand and natural gas prices.

14 SOLAR ENERGY↗
DOE OSTI2017

Statistical Hypothesis Testing using CNN Features for Synthesis of Adversarial Counterexamples to Human and Object Detection Vision Systems

Validating the correctness of human detection vision systems is crucial for safety applications such as pedestrian collision avoidance in autonomous vehicles. The enormous space of possible inputs to such an intelligent system makes it difficult to design test cases for such systems. In this report, we present our tool MAYA that uses an error model derived from a convolutional neural network (CNN) to explore the space of images similar to a given input image, and then tests the correctness of a given human or object detection system on such perturbed images. We demonstrate the capability of our tool on the pre-trained Histogram-of-Oriented-Gradients (HOG) human detection algorithm implemented in the popular OpenCV toolset and the Caffe object detection system pre-trained on the ImageNet benchmark. Our tool may serve as a testing resource for the designers of intelligent human and object detection systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

US Cosmic Visions: New Ideas in Dark Matter 2017 : Community Report

This white paper summarizes the workshop “U.S. Cosmic Visions: New Ideas in Dark Matter” held at University of Maryland from March 23-25. The flagships of the US Dark Matter search program are the G2 experiments ADMX, LZ, and SuperCDMS, which will cover well-motivated axion and WIMP dark matter over a range of masses. The workshop assumes that a complete exploration of this parameter space remains the highest priority of the dark matter community, and focuses instead on the science case for additional new small-scale projects in dark matter science that complement the G2 program (and other ongoing projects worldwide). It therefore concentrates on exploring distinct, well-motivated parameter space that will not be covered by the existing program; on surveying ideas for such projects (i.e. projects costing ~$10M or less); and on placing these ideas in a global context. The workshop included over 100 presentations of new ideas, proposals and recent science and R&D results from the US and international scientific community.

79 ASTRONOMY AND ASTROPHYSICS↗