Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “generative AI”

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 415 records · Page 23

Rapid AI-based Dissection of Ashes using Raman and XRF Spectroscopy (RADAR-X)

Waste-to-energy (WTE) facilities incinerate ~35 million tons of municipal solid waste annually in the United States. The incineration process reduces the mass and the volume of the waste fraction by over 75 and 95%, respectively. The fraction left after incineration remains as ash residues and is referred to as WTE ash, compromising of bottom and fly ash. In the United States, ~10 million tons of WTE ashes are generated annually and predominantly landfilled because of the lack of secondary end-use pathways. This incurs a significant financial burden (landfilling costs) on U.S. WTE facilities and also results in the loss of materials to the landfill. The primary objective of this research is to understand better the elemental and mineralogical composition of WTE ashes from diverse origins and find composition dependent upcycling pathways for diverting these ashes from landfills. This primary objective was addressed through three research tasks –(Task I) An AI-led Multi-Modal Approach for Compositional Analysis, (Task II) Developing a Dissolution-Based Test for Real Time Analysis, and (Task III) Establishing composition-dependent end uses. The chemical composition of WTE ash is dependent on two factors, i.e., the input waste composition and the operational parameters of a WTE facility (combustion conditions). Amongst these two factors, the input waste composition will likely show spatial and temporal variation. As a result, the chemical composition of WTE ash will also fluctuate. To understand the spatial and temporal variation in WTE ash composition, in Task I, we collected 128 ash samples (62 bottom ash and 66 fly ash samples) from 11 WTE facilities located in 11 U.S. states and characterized them via X-ray Fluorescence, powder X-ray Diffraction, and Raman Spectroscopy. The findings from this extensive characterization work indicated that the key elements in WTE fly ashes are Ca, Cl (greater than 10 wt. %), Si, S, K, Zn ( between 1 and 10 wt. %), Mg, Al, P, Ti, Fe, Cu, Br, and Pb (between 0.1 and 1 wt. %). Similarly, the key elements in WTE bottom ash fraction finer than 45μm are Ca (greater than 10 wt. %), Mg, Al, Si, S, Cl, K, Ti, Fe, Zn (between 1 and 10 wt. %), P, V, Cr, Mn, Cu, Br, and Pb (between 0.1 and 1 wt. %). Here, we note that the dominant fraction of WTE bottom ash is the coarse fraction. The coarse WTE bottom ash fraction (rich in silicon) was not characterized in this study because of excessive grinding requirements and their unsuitability as a supplementary cementitious material due to their coarse nature. The elements in WTE bottom ashes are present as calcite, anhydrite, vaterite, hydroxyapatite, quartz, bassanite, gehlenite, akermanite, hydrocalumite, and portlandite. Similarly, the mineralogical species present in WTE fly ashes are calcium chloride hydroxide, halite, calcite, anhydrite, sylvite, hydrocalumite, vaterite, hannebachite, quartz, and bassanite. Temporal variation in ash composition may also result in significant fluctuations in chemical compositions. Therefore, a WTE facility may need to monitor the ash composition (elemental and mineralogical composition) in real time. In Task I, we evaluated the possibility of using a portable X-ray fluorescence (XRF) spectrometer to monitor the elemental composition in real-time. Specifically, we collected XRF data on identical specimens via a portable XRF spectrometer (low-end) and a lab-based benchtop XRF spectrometer (high-end). The collected data was used to train an A.I. algorithm (portable XRF data as an input and benchtop XRF data as an output) to predict accurate elemental composition using portable XRF data. Finally, we developed a 2-minute photobleaching protocol to monitor the mineralogical characteristics of WTE ashes via Raman spectroscopy. Overall, the activities in Task I improved our understanding of ash composition and developed techniques to monitor elemental and mineralogical composition in near real-time. Based on the findings of Task I, we find that WTE ashes exhibit wide variability in mineralogy. For ICP-based elemental analysis, all the mineralogical species in WTE ashes must be brought into solution. This is traditionally accomplished with acid digestion using a combination of multiple acids. However, acid digestion with multiple acids is time-consuming and often fails to ensure complete digestion of the ash matrix. To address this limitation, in Task II, we developed an alkali-fusion-based digestion protocol using a combination of lithium tetraborate, lithium metaborate, and their combinations as possible alkali fluxes for digesting WTE ashes entirely and rapidly. The validity of the developed method was evaluated on two standard ash specimens, i.e., SRM 1633c coal fly ash and BCR-176R incineration fly ash specimen. The findings suggest that the developed protocol can ensure complete digestion of elements such as Al, Ba, Ca, Cr, Cu, Mg, Mn, P, Sr, V, Zn, Be, K, and rare earth elements. The recent changes in the energy market towards renewables and increased metal recycling have resulted in reduced supplies of supplementary cementitious materials (coal fly ash and slag). Therefore, in Task III, we evaluated the possibility of employing WTE ashes as SCMs. As the chemical composition of WTE ashes varies temporally (on an hourly basis), there was also a need to develop tests that can evaluate the suitability of material to act as supplementary cementitious material rapidly, i.e., in a few minutes. Therefore, in Task II, we also developed a rapid test to assess the suitability of a material to act as an SCM in 5 minutes. This represents a significant advance over the state-of-the-art R 3 test, which takes ~144 hours. This test was initially validated on amorphous aluminosilicates, such as calcined clays, and could be extended to evaluate other industrial by-products, such as WTE ashes. In Task III, we evaluated the possibility of employing WTE ashes for two applications, i.e., as an SCM and a lime substitute for clay stabilization. The findings from Task I indicated that WTE ashes are enriched in chlorine and, therefore, cannot be used directly as an SCM due to corrosion-related risks and altered hydration kinetics. Accordingly, we developed an ash treatment protocol to reduce the solubility of chlorine-containing species in WTE ashes. The developed treatment protocol also immobilized lead in certain mineral forms. As a result of the treatment, WTE ashes can be used as SCMs without any corrosion or heavy metal leaching concerns. The second application examined in this study was clay stabilization. WTE ashes are calcium-rich and can be an adequate lime replacement for clay stabilization. Our findings reveal that the sum of the concentrations of Ca(OH) 2 and CaClOH controls the clay stabilization capability of WTE ashes. In summary, in this work, we evaluated the elemental and mineralogical characteristics of U.S. WTE ashes from diverse origins and developed tests to evaluate the chemical characteristics of these ashes in real time through a portable XRF and a benchtop Raman spectrometer. Based on the chemical characteristics of these ashes, we developed an ash treatment process to enable the use of WTE ashes as an SCM and also evaluated the possibility of employing these ashes for clay stabilization. Overall, the findings from this work enables the diversion of WTE ashes from landfills for multiple end-uses, i.e., as an SCM or a lime substitute for clay stabilization.

36 MATERIALS SCIENCE↗

Networks and interfaces as catalysts for polymer materials innovation

Autonomous experimental systems offer a compelling glimpse into a future where closed-loop, iterative cycles—performed by machines and guided by artificial intelligence (AI) and machine learning (ML)—play a foundational role in materials research and development. This perspective draws attention to the roles of networks and interfaces—of and between humans and machines—for the purpose of generating knowledge and accelerating innovation. Polymers, a class of materials with massive global impact, present a unique opportunity for the application of informatics and automation to pressing societal challenges. To develop these networks and interfaces in polymer science, the Community Resource for Innovation in Polymer Technology (CRIPT)—a polymer data ecosystem based on novel polymer data model, representation, search, and visualization technologies—is introduced. The ongoing co-design efforts engage stakeholders in industry, academia, and government to uncover rapidly actionable, high-impact opportunities to build networks, bridge interfaces, and catalyze innovation in polymer technology.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Integration of Simulated and Real Distributed Acoustic Sensor Measurements to Develop AI Enhanced Intelligent Sensing Operations

Distributed acoustic sensors (DAS) have shown remarkable success in monitoring critical civil, energy and transportation assets over long distances and under harsh environmental conditions. Integration of artificial intelligence (AI) technologies with DAS can extend their operational capabilities from conventional tasks (e.g., vibration frequency detection) to more advanced intelligent tasks like anomaly detection. However, these AI technologies often require large labeled/ annotated DAS measurements datasets of the assets in both normal and anomalous operating conditions. This data acquisition task can be both time consuming and costly. This work develops a generative adversarial network based unsupervised domain adaptation framework to build intelligent DAS operational capabilities.

Venketeswaran, Abhishek↗

Artificial Intelligence and Machine Learning Applications in Modern Power Systems

Machine learning (ML) and artificial intelligence (AI) algorithms offer valuable tools for the analysis and interpretation of large datasets. These tools have the capability to uncover insights that may not be readily apparent within these datasets. In recent years, the integration of ML and AI has become increasingly prevalent in various applications within the power system domain. One of the earliest instances of machine learning in power systems can be traced back to demand forecasting, where artificial neural networks were employed for short-term load forecasting. In contemporary power systems, an abundance of high-resolution geospatial and temporal data is generated at various time intervals, ranging from sub-seconds (Phasor Measurement Units or PMUs) to seconds (Supervisory Control and Data Acquisition or SCADA), minutes (Process Information or PI), and extending to days, months, and years. These datasets contain valuable information concerning system reliability and performance. This information holds the potential to offer critical insights into system operations, as well as solutions for predicting and mitigating contingencies to prevent cascading outages. Despite the immense power of machine learning tools, system operators, planners, and utilities often exhibit hesitancy in fully embracing AI-enabled system operations and planning. This cautious approach persists, even as numerous diverse applications of machine learning continue to emerge in the realm of power systems. In this chapter, our focus will delve deep into ML and AI applications tailored for power systems. These applications aim to furnish system operators with enhanced situational awareness and augment their decision-making capabilities, especially during challenging operating conditions. Specific areas of interest encompass root cause analyses of electricity market datasets and the strategic selection of representative samples from vast power system databases for training ML/AI models. Finally, the chapter will conclude with a short discussion on the future of ML/AI in power systems and possible directions that the industry is moving towards.

power system applications, machine learning (ML), ↗

A human-in-the-loop explanation framework for morphologically transparent AI predictions from whole-slide images

Deep learning models enable the prediction of clinical endpoints from whole-slide images (WSIs), but many such models function as “black boxes”, lacking transparency about whether and which histomorphological patterns drive their predictions, hindering interpretability and clinical adoption. Here we propose a human-in-the-loop explanation framework, MorphoXAI, which provides both local and global interpretability for deep learning models by incorporating human-expert interpretations. At the global level, it reveals the histomorphological patterns on which the model consistently relies to distinguish between classes of WSIs, as well as the patterns associated with confusion between classes. At the local level, it indicates which of these patterns are used in the prediction of an individual WSI and which regions within the slide correspond to such patterns. We validated our method across multiple deep learning–based WSI analysis tasks spanning different tissue types. The results show that our framework generates explanations that accurately reflect the histomorphology underlying the model’s predictions at both global and local levels. For interpretability and clinical utility in diagnostic contexts, human evaluation results showed that our explanations were easy to interpret, rich in diagnostic features, and directly helpful for diagnostic decision-making, thereby enhancing pathologist-AI collaboration. Our work highlights that unifying global and local explanations and grounding them in expert-interpreted morphology enhances the interpretability and verifiability of deep learning models, thereby facilitating the transparent deployment of such models in clinical practice.

Lou, Peiliang↗

SDYN-GANs: Adversarial learning methods for multistep generative models for general order stochastic dynamics

We introduce adversarial learning methods for data-driven generative modeling of dynamics of nth-order stochastic systems. Our approach builds on Generative Adversarial Networks (GANs) with generative model classes based on stable m-step stochastic numerical integrators. From observations of trajectory samples, we introduce methods for learning long-time predictors and stable representations of the dynamics. Our approaches use discriminators based on Maximum Mean Discrepancy (MMD), training protocols using both conditional and marginal distributions, and methods for learning dynamic responses over different time-scales. We show how our approaches can be used for modeling physical systems to learn force-laws, damping coefficients, and noise-related parameters. Our adversarial learning approaches provide methods for obtaining stable generative models for dynamic tasks including long-time prediction and developing simulations for stochastic systems.

• Artificial intelligence (AI) / machine learning ↗

Cellular alignment and fusion: Quantifying the effect of macrophages and fibroblasts on myoblast terminal differentiation

Highlights: • We present a method for cellular alignment calculation during routine cell culture. • Application revealed that macrophages promote myoblast alignment, but delay fusion. • Furthermore, fibroblasts regulate myoblast fusion, but not alignment. • Effect of macrophages on alignment is not altered by co-culture with fibroblasts. Successful skeletal muscle wound repair requires the alignment and fusion of myoblasts to generate multinucleated myofibers. In vitro, the accurate quantification of cellular alignment remains a challenge. Here we present the application of ImageJ and ct-FIRE to quantify muscle cell orientation by means of an alignment index (AI). Our optimised method, which does not require programming skills, allows the alignment of myoblasts in vitro to be determined independently of a predefined reference point. Using this method, we demonstrate that co-culture of myoblasts with macrophages, but not fibroblasts, promotes myoblast alignment in a cell density-dependent manner. Interestingly, myoblast fusion was significantly decreased in response to co-culture with macrophages, while the effect of fibroblasts on fusion was density-dependent. At lower numbers, fibroblasts significantly increased myoblast fusion, whereas at higher numbers a significant decrease was observed. Finally, triple co-culture revealed that the effect of macrophages on myoblast alignment and fusion is unaltered by the additional presence of fibroblasts. Application of our optimised method has therefore revealed quantitative differences in the roles of macrophages versus fibroblasts during alignment and fusion: while successful myoblast alignment is promoted by increasing macrophage numbers, regenerative fusion coincides with a decreasing macrophage population and initial rise in fibroblast numbers.

60 APPLIED LIFE SCIENCES↗

Laser-Doppler vibrating tube densimeter for measurements at high temperatures and pressures

A laser-Doppler vibrometer was used to measure the vibration of a vibrating tube densimeter for measuring P-V-T data at high temperatures and pressures. The apparatus developed allowed the control of the residence time of the sample so that decomposition at high temperatures could be minimized. A function generator and piezoelectric crystal was used to excite the U-shaped tube in one of its normal modes of vibration. Densities of methanol-water mixtures are reported for at 673 K and 40 MPa with an uncertainty of 0.009 g/cm{sup 3}.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Industry Level Feasibility of LiDAR Data into Fire Modeling Using Fire Risk Investigation in 3D (FRI3D)

Many evaluation, assessment, and modeling tasks at nuclear power plants require spatial information this often requires physical visits to locations within the facility because the 2D or 3D schematics and current models do not contain enough detail or do not capture as-built and real-world conditions. These visits require extensive manual labor for not only the requesting party, but support groups such as security. LIDAR mapping is trying to solve that problem by providing very detailed 3D models for low costs. However, the use of these models can be very limited because either component reference information is missing and too costly to add, or there is no way to extract specific spatial data needed for other tools. This report presents Idaho National Lab's work with Environmental Intellect (Ei) covering two main efforts. First, to reduce the effort of "tagging" data in large 3D models. By using both existing plant database information, and artificial intelligence (AI) to find and read equipment labels. This research explores the ability of to provide a simple way for the user to tag items and verify plant data, capturing both the speed of AI and human verification. The second part of the work is the development of an interface for importing pieces needed for Modeling & Simulation. Analysis work such as that for fire, flood, or physical security all require spatial or 3D models in various levels of detail. This interface will allow for the retrieval of item location or boundaries, enabling the auto generation of models for varying tools. The application program interface (API) of the fire risk investigation in 3D (FRI3D) was used to test feasibility of exporting the LiDAR tagged spatial information. Outcomes from this work provide preliminary data to determine if the tools and methods could provide substantial industry benefit if fully matured.

97 MATHEMATICS AND COMPUTING↗

A Multi-Objective Bayesian Optimized Human Assessed Multi-Target Generated Spectral Recommender System for Rapid Pareto Discoveries of Material Properties

Optimization for different tasks like material characterization, synthesis, and functional properties for desired applications over multi-dimensional control parameter and function spaces need a rapid strategic search through active learning. However, in all cases prior to optimization, the target material properties are assumed known and fixed, which mostly deviates from real-world scenarios in material synthesis. This can be critical for running expensive experiments on new materials, when the experimental results are fuzzy for any scientific outcomes due to improper target setting, ultimately wasting time and cost. The failure rate and cost are even higher over exploring on multi-target space, where we want to learn the pareto among multiple properties, to jointly optimize during material synthesis for desired applications. To address the challenge, here we introduce the human-operator attempt flexibility in the active learning based automated experiment framework, with generating multiple human assessed targets through a voting-based recommender system during real-time microscope measurements over the large material image space, sequentially learn/update multiple desired targets through a weighting system, and adaptively search in multiple material properties functional space for non-dominated pareto discoveries to maximize the custom structural similarity based acquisition function. We term this a multi-objective Bayesian optimized human assessed multi-target generated spectral recommender systems (MOBO-HAM-SRS). The approach has been demonstrated to peizoresponse force spectroscopy of a ferroelectric thin film, exploring with different kernels and acquisition functions. This work shows an advancement towards human-AI collaborated automated experiments, steering optimization trajectories through human overpowering AI at the early stage when uncertainty is high and AI overpowering human at the later stage with rapid exploration towards optimal goal, following human-assessed multiple targets properties.

Biswas, Arpan↗

Transforming Energy through Computational Science: Computing for Clean Energy

This fact sheet discusses the opportunity space for NREL's computational science capabilities to address national clean energy objectives. Achieving a carbon-free power sector by 2035 as a step towards a decarbonized U.S. energy economy in 2050 will require major advances in power generation, autonomous energy systems, transportation, and buildings/communities. Development of integrated modeling approaches for complex energy systems will be essential for deployment. Success requires developments in optimization and control theory, complemented by machine learning (ML) and artificial intelligence (AI), all of which in turn need targeted investments in breadth and scale of computing. This document defines an opportunity space where: embracing computing can link established research and development (R&D) to a decarbonization agenda; pursuing emerging approaches can accelerate the pace of technology advancement across the portfolio; and leading by example could reduce the carbon footprint of computing worldwide.

advanced computing↗

AutoTG: Reinforcement Learning-Based Symbolic Optimization for AI-Assisted Power Converter Design

Power converters are pervasive in modern electronic component design. They can be found in all electronic devices from household appliances and cellphone chargers to vehicles. Currently, designing new circuit topologies is hard because it requires human expertise based on experience and is difficult to automate. However, artificial-intelligence-assisted design can significantly facilitate the development of new power converters and/or improve the final result. Intelligently designed highly efficient power converters can have a significant effect on many important attributes, such as power efficiency, layout size, cost, heat dissemination, energy requirements, etc. We propose Autonomous Topology Generator (AutoTG), a reinforcement-learning-based framework that generates power converter topology candidates based on user specifications, optimized for user preferences. By modeling power converter design as a symbolic optimization problem, we sequentially sample components in an autoregressive manner until new topologies are formed, providing both the topology specification and the sizing (magnitude of each component parameter) of the proposed power converter. Here, we provide an empirical evaluation and show that AutoTG is able to generate varied high-efficiency topologies within component restrictions based on user input and show that previously unknown topologies can be found for further evaluation.

(AI)-based design↗

AI Denoising to Accelerate Detector Simulation

Detector simulation is critical to experimental HEP; however this simulation (commonly done through toolkits such as Geant4) is computationally intensive. Performance can be improved somewhat through technical optimization, but more is needed. Using machine learning (ML) to accelerate simulation is a promising field, however efforts to use generative adversarial networks (GANs) or optimized autoencoders have faced issues. Using convolutional neural networks (CNNs) for denoising has been successful in non-HEP applications such as image processing. This poster investigates the efficacy of using CNNs to denoise Geant4 simulations. This could increase the accuracy of simulations performed under settings designed to increase computational efficiency.

Franklin, Lena↗

Gaia: An AI-enabled genomic context–aware platform for protein sequence annotation

Protein sequence similarity search is fundamental to biology research, but current methods are typically not able to consider crucial genomic context information indicative of protein function, especially in microbial systems. Here, we present Gaia (Genomic AI Annotator), a sequence annotation platform that enables rapid, context-aware protein sequence search across genomic datasets. Gaia leverages gLM2, a mixed-modality genomic language model trained on both amino acid sequences and their genomic neighborhoods to generate embeddings that integrate sequence-structure-context information. This approach allows for the identification of functionally and/or evolutionarily related genes that are found in conserved genomic contexts, which may be missed by traditional sequence- or structure-based search alone. Gaia enables real-time search of a curated database comprising more than 85 million protein clusters from 131,744 microbial genomes. We compare the homolog retrieval performance of Gaia search against other embedding and alignment-based approaches. We provide Gaia as a web-based, freely available tool.

Jha, Nishant↗

LLM Benchmarking with LLaMA2: Evaluating Code Development Performance Across Multiple Programming Languages

The rapid evolution of large language models (LLMs) has opened new possibilities for automating various tasks in software development. This paper evaluates the capabilities of the LLaMA 2-70B model in automating these tasks for scientific applications written in commonly used programming languages. Using representative test problems, we assess the model's capacity to generate code, documentation, and unit tests, as well as its ability to translate existing code between commonly used programming languages. Our comprehensive analysis evaluates the compilation, runtime behavior, and correctness of the generated and translated code. Additionally, we assess the quality of automatically generated code, documentation, and unit tests. Here, our results indicate that while LLaMA 2-70B frequently generates syntactically correct and functional code for simpler numerical tasks, it encounters substantial difficulties with more complex, parallelized, or distributed computations, requiring considerable manual corrections. We identify key limitations and suggest areas for future improvements to better leverage AI-driven automation in scientific computing workflows.

97 MATHEMATICS AND COMPUTING↗

ECP Report: Update on Proxy Applications and Vendor Interactions

The ExaLearn miniGAN team (Ellis and Rajamanickam) have released miniGAN, a generative adversarial network(GAN) proxy application, through the ECP proxy application suite. miniGAN is the first machine learning proxy application in the suite (note: the ECP CANDLE project did previously release some benchmarks) and models the performance for training generator and discriminator networks. The GAN's generator and discriminator generate plausible 2D/3D maps and identify fake maps, respectively. miniGAN aims to be a proxy application for related applications in cosmology (CosmoFlow, ExaGAN) and wind energy (ExaWind). miniGAN has been developed so that optimized mathematical kernels (e.g., kernels provided by Kokkos Kernels) can be plugged into to the proxy application to explore potential performance improvements. miniGAN has been released as open source software and is available through the ECP proxy application website (https://proxyapps.exascaleproject.ordecp-proxy-appssuite/) and on GitHub (https://github.com/SandiaMLMiniApps/miniGAN). As part of this release, a generator is provided to generate a data set (series of images) that are inputs to the proxy application.

97 MATHEMATICS AND COMPUTING↗

Modeling Framework for Data Center

This chapter highlights the critical need for advanced modeling of data centers due to their rapidly increasing energy consumption and impact on grid reliability. Driven by the demand for AI applications, data centers are projected to consume a significant portion of US energy by 2028, putting stress on an already challenged power grid. The chapter emphasizes the importance of "fast" time-scale models to understand the dynamic interactions between data centers and the grid, especially given the rapid power fluctuations of AI workloads. It outlines a modeling framework that includes both offline and real-time EMT domain simulations, detailing the necessary representations for various components like utility interfaces, transformers, IT loads, UPS, cooling loads, Battery Energy Storage Systems (BESS), generators, protection systems, and higher-level control systems. While standard simulation tools like PSCAD offer basic models, custom development is often required to accurately capture the unique and fast-changing behaviors of modern data centers. The chapter also discusses key metrics and test cases for validating these models, focusing on transient load responses, protection relay coordination, and demand flexibility. Finally, it addresses the challenges of modeling large-scale data centers, such as computational complexity and the trade-off between model fidelity and practicality, suggesting hybrid modeling approaches as a solution. The overarching goal is to create a robust framework that helps assess data center impacts on grid stability, identify vulnerabilities, and inform the development of standards for reliable integration of these large loads into the bulk power system.

25 ENERGY STORAGE↗