Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “machine reasoning”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Evaluating the Effectiveness of Retrieval-Augmented Large Language Models in Scientific Document Reasoning

Despite the dramatic progress in Large Language Model (LLM) development, LLMs often provide seemingly plausible but not factual information, often referred as hallucinations. Retrieval-augmented LLMs provide a non-parametric approach to solve these issues by retrieving relevant information from external data sources and augment the training process. These models helps to trace evidence from an externally provided knowledge base allowing the model predictions to be better interpreted and verified. In this work, we critically evaluate these models in their ability to perform in scientific document reasoning tasks. To this end, we tuned multiple such model variants with science-focused instructions and evaluated them on a scientific document reasoning benchmark for the usefulness of the retrieved document passages. Our findings suggest that models justify predictions in science tasks with fabricated evidence and leveraging scientific corpus as pretraining data does not alleviate the risk of evidence fabrication.

• Artificial intelligence (AI) / machine learning ↗
DOE OSTI2024

Measurement of seasonal variations of the NSLS-II circumference and energy

Routine beam-based characterization of the NSLS-II storage ring indicated probable seasonal variations of the ring circumference and energy. We arranged regular direct measurements of the ring circumference during 2023 and compared the results with the archived data of the magnet and undulator settings. A clear correlation between the ring circumference and energy was observed. These seasonal variations of the ring energy are found to be the main reason for regular lattice corrections required to keep the machine performance stable for user operations. Finally, the measurement techniques and results are discussed in this article.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Materials structure–property factorization for identification of synergistic phase interactions in complex solar fuels photoanodes

Abstract Properties can be tailored by tuning composition in high-order composition spaces. For spaces with complex phase behavior, modeling the properties as a function of composition and phase distribution remains a formidable challenge. We present materials structure–property factorization (MSPF) as an approach to automate modeling of such data and identify synergistic phase interactions. MSPF is an interpretable machine learning algorithm that couples phase mapping via Deep Reasoning Networks (DRNets) to matrix factorization-based modeling of the representative properties of each phase in a dataset. MSPF is demonstrated for Bi–Cu–V oxide photoanodes for solar fuel generation, which contains 25 different phase combinations and correspondingly exhibits complex composition-structure-photoactivity relationships. Comparing the measured photoactivity to a learned model for non-interacting phases, synergistic phase interactions are identified to guide further photoactivity optimization and understanding. MSPF identifies synergistic interactions of a BiVO 4 -like phase with both Cu 2 V 2 O 7 -like and CuV 2 O 6 -like phases, creating avenues for understanding complex photoelectrocatalysts.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Reweighting configurations generated by transferable, machine learned models for protein sidechain backmapping

Multiscale modeling requires the linking of models at different levels of detail, with the goal of gaining accelerations from lower fidelity models while recovering fine details from higher resolution models. Communication across resolutions is particularly important in modeling soft matter, where tight couplings exist between molecular-level details and mesoscale structures. While multiscale modeling of biomolecules has become a critical component in exploring their structure and self-assembly, backmapping from coarse-grained to fine-grained, or atomistic, representations presents a challenge, despite recent advances through machine learning. A major hurdle, especially for strategies utilizing machine learning, is that backmappings can only approximately recover the atomistic ensemble of interest. We demonstrate conditions for which backmapped configurations may be reweighted to exactly recover the desired atomistic ensemble. By training separate decoding models for each sidechain type, we develop an algorithm based on normalizing flows and geometric algebra attention to autoregressively propose backmapped configurations for any protein sequence. Critical for reweighting with modern protein force fields, our trained models include all hydrogen atoms in the backmapping and make probabilities associated with atomistic configurations directly accessible. We also demonstrate, however, that reweighting is extremely challenging despite state-of-the-art performance on recently developed metrics and generation of configurations with low energies in atomistic protein force fields. Through detailed analysis of configurational weights, we show that machine-learned backmappings must not only generate configurations with reasonable energies, but also correctly assign relative probabilities under the generative model. These are broadly important considerations in generative modeling of atomistic molecular configurations.

Monroe, Jacob I. [Univ. of Arkansas, Fayetteville,↗
DOE OSTI2025

Automated and highly parallelized Bayesian optimization scheme for direct drive fusion experiments on OMEGA

Finding the optimal implosion design on existing experimental facilities for inertial confinement fusion requires an exhaustive search of the vast design parameter space. This is infeasible both with experiments and with simulations. Consequently, a large fraction of the experimentally realizable design space remains unexplored, and new design schemes are challenging to optimize in a reasonable time frame. On the OMEGA laser facility, predictive machine learning models have been developed to accurately forecast the result of an experiment using only inexpensive simulations and the large dataset of prior experimental data. However, the full design space remains vast enough to be unassailable with simple optimization techniques. Here we develop an automated and optimally parallel Bayesian optimization algorithm that can entirely optimize the target and pulse shape of a direct-drive ICF implosion under a given design paradigm. We use this algorithm to find a markedly improved design for the performance implosions on OMEGA that is predicted to hydroequivalently scale to ignition at 2.15 MJ.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Report on Pure Mathematics of Machine Learning

We report on the findings conducted by the Pure Mathematics in Machine Learning (PMML) Project under the umbrella of the Mathematics in Artificial Reasoning in Sciences (MARS) Laboratory Initiative at the Pacific Northwest National Laboratory. The findings contained in this report spans three years of research from the project team. The objectives of the PMML project are two-fold: 1. To gain understanding of modern machine learners through empirical research via structured and sound experimental design, and 2. To prove mathematical theories based upon these empirical findings. We present both empirical results along with mathematical theories in this document.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Towards Pareto optimal high entropy hydrides via data-driven materials discovery

The ability to rapidly screen material performance in the vast space of high entropy alloys is of critical importance to efficiently identify optimal hydride candidates for various use cases. Given the prohibitive complexity of first principles simulations and large-scale sampling required to rigorously predict hydrogen equilibrium in these systems, we turn to compositional machine learning models as the most feasible approach to screen on the order of tens of thousands of candidate equimolar high entropy alloys (HEAs). Critically, we show that machine learning models can predict hydride thermodynamics and capacities with reasonable accuracy (e.g. a mean absolute error in desorption enthalpy prediction of ~5 kJ mol H 2 –1 ) and that explainability analyses capture the competing trade-offs that arise from feature interdependence. We can therefore elucidate the multi-dimensional Pareto optimal set of materials, i.e., where two or more competing objective properties can't be simultaneously improved by another material. This provides rapid and efficient down-selection of the highest priority candidates for more time-consuming density functional theory investigations and experimental validation. Various targets were selected from the predicted Pareto front (with saturation capacities approaching two hydrogen per metal and desorption enthalpy less than 60 kJ mol H 2 –1 ) and were experimentally synthesized, characterized, and tested amongst an international collaboration group to validate the proposed novel hydrides. Finally, additional top-predicted candidates are suggested to the community for future synthesis efforts, and we conclude with an outlook on improving the current approach for the next generation of computational HEA hydride discovery efforts.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Applying Cyber-Informed Engineering to Power System Operations

This presentation covers the interaction of the discipline of system operations with the growing body of knowledge around Cyber-Informed Engineering (CIE). First is discussion of a number of fundamental concepts for system operators - organizational division of responsibilities, human and machine cooperation, goals, and priorities. The next section covers the reasons why CIE was developed, what it is, and the key design and operational, and organizational principles of CIE. Finally, some thoughts on how CIE can be applied to power system operations are offered, along with some examples of how an organization can approach applying CIE principles in their particular circumstances.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Unraveling Hydrogen Induced Geochemical Reaction Mechanisms through Coupled Geochemical Modeling and Machine Learning

Underground hydrogen storage (UHS) provides a promising large-scale, long-term energy storage solution. A reasonable recovery of stored hydrogen is critical for a successful storage scheme. However, in subsurface reservoirs hydrogen is subject to active geochemical reactions that might result in hydrogen loss. In this study, we implemented a geochemical modeling approach coupled with an unsupervised machine learning technique called non-negative matrix factorization (NMF) to unravel the complex brine-rock-H 2 geochemical processes responsible for hydrogen losses, with particular focus on sulfate reduction reactions. NMF is applied to modeled mineral evolution and fluid component profiles to retrieve profiles that can be interpreted to more easily assess competing processes. NMF decouples simulated competing equilibrium reactions. This facilitates separation of overlapping reaction profiles from redox processes, dissolution fronts, and secondary precipitation while considering the effects of simulation parameters such as salinity, temperature, and total H 2 pressure. NMF successfully discriminates these competing effects in nonlinear ways, allowing robust interpretation. In addition, NMF reveals subtle coupled mineral associations and reaction fronts that are invisible to conventional model analysis. This integrated approach strengthens the conceptual understanding of complex nonlinear hydrogen-brine-rock interactions and advances geochemical research on UHS systems to resolve complexities in modeled geochemical systems without the need for direct experiments or prior knowledge. Furthermore, this study highlights the efficacy of combining geochemical modeling with machine learning techniques to enhance the interpretability of the intricate geochemical simulation output through deciphering the overlapping reaction path that cannot be achieved only using conventional analysis of geochemical models alone.

08 HYDROGEN↗
DOE OSTI2024

Evaluation of GlassNet for physics-informed machine learning of glass stability and glass-forming ability

Glassy materials form the basis of many modern applications, including nuclear waste immobilization, touch-screen displays, and optical fibers, and also hold great potential for future medical and environmental applications. However, their structural complexity and large composition space make design and optimization challenging for certain applications. Of particular importance for glass processing and design is an estimate of a given composition's glass-forming ability (GFA). However, there remain many open questions regarding the underlying physical mechanisms of glass formation, especially in oxide glasses. It is apparent that a proxy for GFA would be highly useful in glass processing and design, but identifying such a surrogate property has proven itself to be difficult. While glass stability (GS) parameters have historically been used as a GFA surrogate, recent research has demonstrated that most of these parameters are not accurate predictors of the GFA of oxide glasses. Here, in this work, we explore the application of an open-source pre-trained neural network model, GlassNet, that can predict the characteristic temperatures necessary to compute GS with reasonable performance and assess the feasibility of using these physics-informed machine learning (PIML)-predicted GS parameters to estimate GFA. In doing so, we track the uncertainties at each step of the computation—from the original ML prediction errors to the compounding of errors during GS estimation, and finally to the final estimation of GFA. While GlassNet exhibits reasonable accuracy on all individual properties, we observe a large compounding of error in the combination of these individual predictions for the PIML prediction of GS, finding that random forest models offer similar accuracy to GlassNet. We also break down the performance of GlassNet on different glass families and find that the error in GS prediction is correlated with the error in crystallization peak temperature prediction. Lastly, we utilize this finding to assess the relationship between top-performing GS parameters and GFA for two ternary glass systems: sodium borosilicate and sodium iron phosphate glasses. We conclude that to obtain true ML predictive capability of GFA, significantly more data needs to be collected.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Predicting the superconducting critical temperature in transition metal carbides and nitrides using machine learning

Transition metal carbides and nitrides have unique mechanical and chemical characteristics. At low temperatures many of them also exhibit superconductivity, which can be controlled by substitutions into both the transition metal and carbon/nitrogen sites. To investigate the factors governing the superconducting state, we apply machine learning methods. Here we collected a dataset containing 147 materials, which was used to create a pipeline for predicting their superconducting critical temperature. When this pipeline is applied to a randomly selected test set, it shows a good performance, with of 0.82 and RMSE of 1.9 K. To explore the limits of the machine learning approach, we also use it to predict entire substitution series within the dataset. This represents a realistic test for the predictive models, which can be extremely useful when applied to new substitutions in materials systems. The performance of the pipeline in this case is much more uneven, with good predictions for some series, while for others the model shows minimal predictive power. We discuss possible reasons for these results, as well as methods to estimate the performance of machine learning on new substitution series.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Machine Learning a Simple Interpretable Short-Range Potential for Silica

A wide array of models, spanning from computationally expensive ab initio methods to a spectrum of force-field approaches, have been developed and employed to probe silica polymorphs and understand growth processes and atomic-level dynamical transitions in silica. However, the quest for a model capable of making accurate predictions with high computational efficiency for various silica polymorphs is still ongoing. Recent developments in short-range machine-learned models, such as GAP and NNPScan, have shown promise in providing reasonable descriptions of silica, but their computational cost remains high compared to force fields such as BKS which are based on simple interpretable functional forms. Here, in this study, we build on the recent success of our reinforcement learning (RL) workflow to derive a new set of optimal parameters for a promising short-range BKS-based model proposed by Soules. We use RL to navigate the eight-dimensional parameter space of the Soules potential using an experimental training data set that includes both local and global structural features from approximately 21 experimentally realized silica polymorphs, including high density phases and porous zeolites. We compare the performance of our machine-learned ML-Soules model with other high quality models including our recent machine-learned parametrization of BKS (ML-BKS), a machine-learned potential (GAP), as well as predictions of ab initio calculations with the highly fidelity SCAN functional. The ML-Soules accurately captures the relative energetic ordering of various polymorphs as well as their structural features at a significantly reduced computational expense. The ML-Soules model also reasonably captures the structure, density, and elastic constants of quartz, as well as metastable silica polymorphs. We further discuss the limitations of the Soules functional form and propose potential enhancements, including the incorporation of additional three-body terms and/or the utilization of different short-ranged functional forms to achieve greater accuracy for both global and local features in the modeling of silica while retaining low computational cost.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

How to see hidden patterns in metamaterials with interpretable machine learning

Machine learning models can assist with metamaterials design by approximating computationally expensive simulators or solving inverse design problems. However, past work has usually relied on black box deep neural networks, whose reasoning processes are opaque and require enormous datasets that are expensive to obtain. Here, in this work, we develop two novel machine learning approaches to metamaterials discovery that have neither of these disadvantages. These approaches, called shape-frequency features and unit-cell templates, can discover 2D metamaterials with user-specified frequency band gaps. Our approaches provide logical rule-based conditions on metamaterial unit-cells that allow for interpretable reasoning processes, and generalize well across design spaces of different resolutions. The templates also provide design flexibility where users can almost freely design the fine resolution features of a unit-cell without affecting the user’s desired band gap.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Machine learning to predict biomass sorghum yields under future climate scenarios

Crop yield modeling is critical in the design of national strategies for agricultural production, particularly in the context of a changing climate. Forecasting yields of bioenergy crops at fine spatial resolutions can help to evaluate near-term and long-term pathways for scaling up bio-based fuel and chemical production, and for understanding the impacts of abiotic stressors such as severe droughts and temperature extremes on potential biomass supply. In this work we used a large dataset of 28,364 Sorghum bicolor yield samples (uniquely identified by county and year of observation), environmental variables, and multiple approaches to analyze historical trends in sorghum productivity across the USA. We selected the most accurate machine learning approach (a variation of the random forest approach) to predict future trends in sorghum yields under four greenhouse gas (GHG) emission scenarios and two irrigation regimes. We identified irrigation practices, vapor pressure deficit, and time (a proxy for technological improvement) as the most important predictors of sorghum productivity. Our results showed a decreasing trend of sorghum yields over future years (on average 2.7% from 2018 to 2099), with greater decline under a high GHG emissions scenario (3.8%) and in the absence of irrigation (4.6%). Geographically, we observed the steepest predicted declines in the Great Lakes (8.2%), Upper Midwest (7.5%), and Heartland (6.7%) regions. Our study demonstrates the use of machine learning to identify environmental controllers of sorghum biomass yield and predict yields with reasonable accuracy. These results can inform the development of more realistic biomass supply projections for bioenergy if sorghum production is scaled up. (c) 2020 Society of Chemical Industry and John Wiley & Sons, Ltd

09 BIOMASS FUELS↗
DOE OSTI2023

A Prediction Method for Catchment Efficiency Loss due to Coaxial Nozzle Wear in Powder Fed Directed Energy Deposition Systems

Abstract Powder fed laser directed energy deposition systems show significant promise and their widespread adoption could greatly advance manufacturing technology; however, a significant amount of gas atomized metal powder is wasted during material deposition. As contaminated powder must be collected and disposed of as hazardous waste, this process is often criticized for its low catchment efficiency. This presents potential implementation and scaling issues, as widespread use of this technology would directly increase the amount of contaminated powder in need of safe disposal. A direct contribution to this waste issue was identified in past work as wearing exterior and interior surfaces on a replaceable nozzle component within the additive head of such a system integrated into a hybrid manufacturing machine tool. A computational fluid dynamics simulation is developed specifically for use within the manufacturing process by predicting catchment efficiency losses before a deposition operation is performed. The model is robust enough to calculate a catchment efficiency estimate within a reasonable transition time before deposition and is designed to replace visual estimation of nozzle deterioration by the machine operator. Both of two previously identified wear modes are represented in the simulation. Their effects on powder stream geometry are discussed in detail and the results are compared with in-situ measurements of powder stream geometry and concentrations. The resulting work contributes to the growing understanding of using prediction and prevention methods on catchment efficiency related part defects within these systems.

DeWitte, Lisa↗
DOE OSTI2025

Prediction of Silicon Content in a Blast Furnace via Machine Learning: A Comprehensive Processing and Modeling Pipeline

Silicon content plays an important role in determining the operational efficiency of blast furnaces (BFs) and their downstream processes in integrated steelmaking; however, existing sampling methods and first-principles models are somewhat limited in their capability and flexibility. Current data-based prediction models primarily rely on a limited set of manually selected furnace parameters. Additionally, different BFs present a diverse set of operating parameters and state variables that are known to directly influence the hot metal’s silicon content, such as fuel injection, blast temperature, and raw material charge composition, among other process variables that have their own impacts. The expansiveness of the parameter set adds complexity to parameter selection and processing. This highlights the need for a comprehensive methodology to integrate and select from all relevant parameters for accurate silicon content prediction. Providing accurate silicon content predictions would enable operators to adjust furnace conditions dynamically, improving safety and reducing economic risk. To address these issues, a two-stage approach is proposed. First, a generalized data processing scheme is proposed to accommodate diverse furnace parameters. Second, a robust modeling pipeline is used to establish a machine learning (ML) model capable of predicting hot metal silicon content with reasonable accuracy. The method employed herein predicted the average Si content of the upcoming furnace cast with an accuracy of 91% among 200 target predictions for a specific furnace provisioned by the XGBoost model. This prediction is achieved using only the past shift’s operating conditions, which should be available in real time. This performance provides a strong baseline for the modeling approach with potential for further improvement through provision of real-time features.

Chemistry↗
DOE OSTI2021

Robustifying Links To Combat Reference Rot

Links to web resources frequently break, and linked content can change at unpredictable rates. These dynamics of the Web are detrimental when references to web resources provide evidence or supporting information. In this paper, we highlight the significance of reference rot, provide an overview of existing techniques and their characteristics to address it, and introduce our Robust Links approach, including its web service and underlying API. Robustifying links offers a proactive, uniform, and machine-actionable way to combat reference rot. In addition, we discuss our reasoning and approach aimed at keeping the approach functional for the long term. To showcase our approach, we have robustified all links in this article.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2021

Beyond Visual Analytics: Human-Machine Teaming for AI-Driven Data Sensemaking

"Detect the expected, discover the unexpected" was the founding principle of the field of visual analytics. This mantra implies that human stakeholders, like a domain expert or data analyst, could leverage visual analytics techniques to seek answers to known unknowns and discover unknown unknowns in the course of the data sensemaking process. We argue that in the era of AI-driven automation, we need to recalibrate the roles of humans and machines (e.g., a machine learning model) as teammates. We posit that by realizing human-machine teams as a stakeholder unit, we can better achieve the best of both worlds: automation transparency and human reasoning efficacy. However, this also increases the burden on analysts and domain experts towards performing more cognitively demanding tasks than what they are used to. In this paper, we reflect on the complementary roles in a human-machine team through the lens of cognitive psychology and map them to existing and emerging research in the visual analytics community. We discuss open questions and challenges around the nature of human agency and analyze the shared responsibilities in human-machine teams.

Sensemaking, human-machine teaming, agency, artifi↗