Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “explainable”

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 847 records · Page 47
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

A primer on artificial intelligence in plant digital phenomics: embarking on the data to insights journey

Artificial intelligence (AI) has emerged as a fundamental component of global agricultural research that is poised to impact on many aspects of plant science. In digital phenomics, AI is capable of learning intricate structure and patterns in large datasets. We provide a perspective and primer on AI applications to phenome research. We propose a novel human-centric explainable AI (X-AI) system architecture consisting of data architecture, technology infrastructure, and AI architecture design. We clarify the difference between post hoc models and 'interpretable by design' models. We include guidance for effectively using an interpretable by design model in phenomic analysis. We also provide directions to sources of tools and resources for making data analytics increasingly accessible. In conclusion, this primer is accompanied by an interactive online tutorial.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Evaluation of Stratocumulus Evolution Under Contrasting Temperature Advections in CESM2 Through a Lagrangian Framework

Abstract This study leveraged a Lagrangian framework to examine the evolution of stratocumulus clouds under cold and warm advections (CADV and WADV) in the Community Earth System Model 2 (CESM2) against observations. We found that CESM2 simulates a too rapid decline in low‐cloud fraction (LCF) and cloud liquid water path (CLWP) under CADV conditions, while it better aligns closely with observed LCF under WADV conditions but overestimates the increase in CLWP. Employing an explainable machine learning approach, we found that too rapid decreases in LCF and CLWP under CADV conditions are related to overestimated drying effects induced by sea surface temperature, whereas the substantial increase in CLWP under WADV conditions is associated with the overestimated moistening effects due to free‐tropospheric moisture and surface winds. Our findings suggest that overestimated drying effects of sea surface temperature on cloud properties might be one of crucial causes for the high equilibrium climate sensitivity in CESM2.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

A Hybrid Biophysical‐Machine Learning Framework for Diurnal Surface Energy Flux Estimation Using Proximal Sensing

Thermal infrared-based remote sensing of surface energy fluxes has traditionally relied on high spatial resolution satellite data with revisit frequencies on the order of weeks. In this study, we evaluate a biophysics-based analytical surface energy balance model for predicting latent energy ( LE ) and sensible heat ( H ) fluxes using proximal sensing observations. The Surface Temperature Initiated Closure (STIC1.2) model has been extensively validated across a wide range of spatial and temporal scales using various satellite-derived thermal infrared data sets. Here we extend this validation by applying STIC at sub-hourly temporal resolution over multiple growing seasons for four distinct agricultural systems. We further develop and evaluate novel STIC variants that incorporate machine learning (ML) techniques to eliminate the need for surface energy balance observations, specifically net radiation and soil heat flux, thereby enhancing model applicability in data-sparse settings. The integration of a ML component to estimate surface available energy is shown to have strong predictive performance for both LE (R 2 = 0.81–0.94) and H (R 2 = 0.46–0.72) across all agricultural systems examined here, demonstrating the potential of hybrid biophysical-machine learning approaches for surface energy balance modeling with minimal data requirements. This study concludes with a novel application of explainable machine learning (exML) to diagnose sources of model error. This exML framework attributes residual prediction errors to both model input variables and environmental drivers not explicitly included in the simulation experiments. This approach provides a new pathway for improving model design and integrating previously overlooked yet influential variables into future model iterations.

evapotranspiration↗
DOE OSTI2024

Pushing the frontiers in climate modelling and analysis with machine learning

Climate modelling and analysis are facing new demands to enhance projections and climate information. Here, in this study, we argue that now is the time to push the frontiers of machine learning beyond state-of-the-art approaches, not only by developing machine-learning-based Earth system models with greater fidelity, but also by providing new capabilities through emulators for extreme event projections with large ensembles, enhanced detection and attribution methods for extreme events, and advanced climate model analysis and benchmarking. Utilizing this potential requires key machine learning challenges to be addressed, in particular generalization, uncertainty quantification, explainable artificial intelligence and causality. This interdisciplinary effort requires bringing together machine learning and climate scientists, while also leveraging the private sector, to accelerate progress towards actionable climate science.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Human Factors Considerations in Artificial Intelligence Applications for Nuclear Power Plants

In recent years, there has been a wave of artificial intelligence (AI) technologies that offer to solve problems from shopping habits to mortgage approvals to critical systems operations. The rapidity of the development of these systems has led to both excitement and apprehension about the roles these systems should play in our modern societies. Furthermore, this paper focuses on the critical infrastructure industry, in general, and nuclear power generation, in particular, and seeks to scrutinize how we can leverage these novel technologies in human-centered ways to maintain or enhance the established high levels of reliability and resilience in these industries. First, we discuss the broader aspects of cognitive systems and activities that are critical to understanding the human-AI space. Then we explore different approaches to explainability in AI and the notions of trust. We then move on to discuss several human factors concepts and methods and how they can support the design of human-AI teams. We then explore recent research related to nuclear power that has been undertaken and evaluate the current industry and regulatory landscapes. Finally, we discuss identified research gaps and recommendations for solving these for the critical infrastructure space.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2025

XSub: Explanation-Driven Adversarial Attack against Blackbox Classifiers via Feature Substitution

Despite its significant benefits in enhancing the transparency and trustworthiness of artificial intelligence (AI) systems, explainable AI (XAI) can unintentionally provide adversaries with insights into blackbox models, increasing their vulnerability to various attacks. In this paper, we develop a novel explanation-driven adversarial attack against blackbox classifiers based on feature substitution, called XSub. The key idea of XSub is to strategically replace important features (identified via XAI) in the original sample with corresponding important features of a different label, thereby increasing the likelihood of the model misclassifying the perturbed sample. XSub only requires a minimal number of queries and can be easily extended to launch backdoor attacks in case the attacker has access to the model's training data. Our evaluation shows that XSub is not only effective and stealthy but also low-cost, showcasing its feasibility across a wide range of AI applications.

adversarial attack↗
DOE OSTI2021

PHYSICS-BASED AUTOMATED REASONING FOR HEALTH MONITORING: SENSOR SET SELECTION

This paper addresses the problem of how to select a sensor set for equipment health monitoring that meets the needs of advanced O&M tasks that target cost reduction. They include maintenance optimization and asset management for the existing fleet and near-autonomous operation as currently envisioned for advanced reactors. The method uses physics-based automated reasoning to provide for a more “explainable” diagnosis. The algorithm is described along with its implementation on a computational cluster. Preliminary results for application to a use case in the current fleet are described.

diagnosis↗
DOE OSTI2022

Technical Basis for Advanced Artificial Intelligence and Machine Learning Adoption in Nuclear Power Plants

The research and development reported here is part of the Technology Enabled Risk-Informed Maintenance Strategy project sponsored by the U.S. Department of Energy’s Light Water Reactor Sustainability program. The primary objective of the research presented in this report is to produce a technical basis for developing explainable and trustable artificial intelligence (AI) and machine learning (ML) technologies. The technical basis will lay the foundation for addressing the technical and regulatory adoption challenges of AI/ML technologies across plant assets and the nuclear industry at scale and to achieve seamless cost-effective automation without compromising plant safety and reliability.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Using iterative random forest to find geospatial environmental and Sociodemographic predictors of suicide attempts

Despite a recent global decrease in suicide rates, death by suicide has increased in the United States. It is therefore imperative to identify the risk factors associated with suicide attempts to combat this growing epidemic. In this study, we aim to identify potential risk factors of suicide attempt using geospatial features in an Artificial intelligence framework. We use iterative Random Forest, an explainable artificial intelligence method, to predict suicide attempts using data from the Million Veteran Program. This cohort incorporated 405,540 patients with 391,409 controls and 14,131 attempts. Our predictive model incorporates multiple climatic features at ZIP-code-level geospatial resolution. We additionally consider demographic features from the American Community Survey as well as the number of firearms and alcohol vendors per 10,000 people to assess the contributions of proximal environment, access to means, and restraint decrease to suicide attempts. In total 1,784 features were included in the predictive model. Our results show that geographic areas with higher concentrations of married males living with spouses are predictive of lower rates of suicide attempts, whereas geographic areas where males are more likely to live alone and to rent housing are predictive of higher rates of suicide attempts. We also identified climatic features that were associated with suicide attempt risk by age group. Additionally, we observed that firearms and alcohol vendors were associated with increased risk for suicide attempts irrespective of the age group examined, but that their effects were small in comparison to the top features. Taken together, our findings highlight the importance of social determinants and environmental factors in understanding suicide risk among veterans.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Yet Another Discriminant Analysis (YADA): A Probabilistic Model for Machine Learning Applications

This paper presents a probabilistic model for various machine learning (ML) applications. While deep learning (DL) has produced state-of-the-art results in many domains, DL models are complex and over-parameterized, which leads to high uncertainty about what the model has learned, as well as its decision process. Further, DL models are not probabilistic, making reasoning about their output challenging. In contrast, the proposed model, referred to as Yet Another Discriminate Analysis(YADA), is less complex than other methods, is based on a mathematically rigorous foundation, and can be utilized for a wide variety of ML tasks including classification, explainability, and uncertainty quantification. YADA is thus competitive in most cases with many state-of-the-art DL models. Ideally, a probabilistic model would represent the full joint probability distribution of its features, but doing so is often computationally expensive and intractable. Hence, many probabilistic models assume that the features are either normally distributed, mutually independent, or both, which can severely limit their performance. YADA is an intermediate model that (1) captures the marginal distributions of each variable and the pairwise correlations between variables and (2) explicitly maps features to the space of multivariate Gaussian variables. Numerous mathematical properties of the YADA model can be derived, thereby improving the theoretic underpinnings of ML. Validation of the model can be statistically verified on new or held-out data using native properties of YADA. However, there are some engineering and practical challenges that we enumerate to make YADA more useful.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Mapping Hsp104 interactions using cross‐linking mass spectrometry

Molecular machines from the AAA+ (ATPases Associated with diverse cellular Activity) superfamily of protein disaggregases play important roles in protein folding, disaggregation and DNA processing. Recent cryo-EM structures of AAA+ molecular machines have uncovered nuanced changes in their conformation that underlie their specialized functions. Structural knowledge of these molecular machines in complex with substrates begins to explain their mechanism of activity. Here, we explore how cross-linking mass spectrometry (XL-MS) can be used to interpret changes in conformation induced by ATP in Hsp104 and how a substrate may interact with Hsp104. We applied a panel of cross-linking reagents to produce cross-linking maps of Hsp104 and interpret our data on previously determined X-ray and cryo-EM structures of Hsp104 from a thermophilic yeast, Calcarisporiella thermophila. We developed an analysis pipeline to differentiate between intra-subunit and inter-subunit contacts within the hexameric homo-oligomer. We identify cross-links that break the asymmetry that is present in Hsp104 in an ATP-hydrolysis competent conformation but is absent in an ATP-hydrolysis-defective mutant. Finally, we identify contacts between Hsp104 and a selected protein (proprotein convertase subtilisin/kexin type 9 PCSK9) to reveal contacts on the central channel of Hsp104 across the length of this protein indicating that we might have trapped interactions consistent with its translocation. Our simple and robust XL-MS-based experiments and methods help interpret how these molecular machines change conformation and bind to other proteins even in the context of homo-oligomeric assemblies enabling coupling state-of-the-art modeling approaches with XL-MS.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2020

Solar Engineering of Thermal Processes, Photovoltaics and Wind, Fifth Edition

The updated Fifth Edition of Solar Engineering of Thermal Processes, Photovoltaics and Wind contains the fundamentals of solar energy and explains how we get energy from the sun. The authors - noted experts on the topic - provide an introduction to the technologies that harvest, store, and deliver solar energy, such as photovoltaics, solar heaters, and cells. The book also explores the applications of solar technologies and shows how they are applied in various sectors of the marketplace.

14 SOLAR ENERGY↗
DOE OSTI2020

The Role of Electron Localization in Covalency and Electrochemical Properties of Lithium-Ion Battery Cathode Materials

Following the fundamental research conducted by J. B. Goodenough, the important role of electron localization induced by elemental substitution is studied. The size and electron negativity of host and substituting ions are two important factors in tuning material properties such as local structure and transition metal (TM) oxygen covalency. However, another factor, electron localization, which is widely studied in catalyst research but largely overlooked for battery materials, deserves systematic studies. A combined investigation using synchrotronbased X-ray spectroscopy and theoretical calculations is carried out on the Li-Co-Mn-O model system in which the substituting cation Mn 4+ , with its 3d 3 electronic structure, is used as a promoter for electron localization. Results indicate that electron localization greatly influences the Co - O bond by making it less covalent, which increases the delithiation voltage. It is also found that during charge/discharge, electron localization tends to make TM K-edge X-ray absorption near edge spectroscopy (XANES) spectra show a more “rigid shift” behavior while electron delocalization makes the XANES exhibit a “shape change.” It clearly explains why the K-edge XANES data of some TM oxides show no “rigid shift” while the nominal valence states changed. This work highlights the importance of electron localization with guidance for XANES interpretation.

25 ENERGY STORAGE↗
DOE OSTI2020

Low Temperature Aggregation Transitions in N3 and Y6 Acceptors Enable Double‐Annealing Method That Yields Hierarchical Morphology and Superior Efficiency in Nonfullerene Organic Solar Cells

Abstract Thermal transition of organic solar cells (OSCs) constituent materials are often insufficiently researched, resulting in trial‐and‐error rather than rational approaches to annealing strategies to improve domain purity to enhance the power conversion efficiency. Despite the potential utility, little is known about the thermal transitions of the modern high‐performance acceptors Y6 and N3. Here, by using an optical method, it is discovered that the acceptor N3 has a clear solid‐state aggregation transition at 82 °C. This unusually low transition not only explains prior optimization protocols, but the transition informs and enables a double‐annealing method that can fine‐tune aggregation and the device morphology. Compared with 16.6% efficiency for PM6:N3:PC 71 BM control devices, higher efficiency of 17.6% is obtained through the improved protocol. Morphology characterization with x‐ray scattering methods reveals the formation of a multilength scale morphology. Moreover, the double‐annealing method is illustrated and easily transferred and validated with Y6‐based devices, using the transition of Y6 at 102 °C. As a result, the PCE improved from 16.0% to 16.8%. Design of high‐performance acceptors with yet lower aggregation transitions might be required for OSCs to successfully transition to low thermal budget industrial processing methods where annealing temperatures on plastic substrates have to be kept low.

Qin, Yunpeng↗
DOE OSTI2020

Near-Zero Power MOF-Based Sensors for NO 2 Detection

Detection and capture of toxic nitrogen oxides (NO x ) is important for emissions control of exhaust gases and general public health. The ability to directly electrically detect trace (0.5–5 ppm) NO 2 by a metal–organic framework (MOF)-74-based sensor at relatively low temperatures (50 °C) is demonstrated via changes in electrical properties of M-MOF-74, M = Co, Mg, Ni. The magnitude of the change is ordered Ni > Co > Mg and explained by each variant's NO 2 adsorption capacity and specific chemical interaction. Ni-MOF-74 provides the highest sensitivity to NO 2 ; a 725× decrease in resistance at 5 ppm NO2 and detection limit <0.5 ppm, levels relevant for industry and public health. Furthermore, the Ni-MOF-74-based sensor is selective to NO 2 over N 2 , SO 2 , and air. Linking this fundamental research with future technologies, the high impedance of MOF-74 enables applications requiring a near-zero power sensor or dosimeter, with the active material drawing <15 pW for a macroscale device 35 mm 2 with 0.8 mg MOF-74. Finally, this represents a 10 4 –10 6 × decrease in power consumption compared to other MOF sensors and demonstrates the potential for MOFs as active components for long-lived, near-zero power chemical sensors in smart industrial systems and the internet of things.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

RKKY Exchange Bias Mediated Ultrafast All-Optical Switching of a Ferromagnet

The discovery of ultrafast helicity-independent all-optical switching (HI-AOS), as well as picosecond all-electrical switching of a ferrimagnet, has inspired the ultrafast spintronics community to explore ultrafast switching of a ferromagnet to achieve practical ultrafast storage and memory devices. Two explored mechanisms of HI-AOS of a ferromagnet in ferromagnet-ferrimagnet heterostructure are: a) exploiting the indirect exchange coupling with and b) injection of non-local spin current originated from a switching ferrimagnet. Here, in this manuscript, exchange mediated HI-AOS of a Ruderman–Kittel–Kasuya–Yosida (RKKY) exchange coupled “[Co/Pt]-multilayers/Pt spacer/CoGd” heterostructure is demonstrated. The authors have measured layer-resolved static magnetic properties, single-shot HI-AOS, and magnetization dynamics of the ferromagnetic Co/Pt multilayers (MLs), that are ferromagnetically or antiferromagnetically coupled with ferrimagnetic CoGd layers. Time-resolved magnetization dynamics reveal a 3.5 ps switching time of the Co/Pt MLs, which is the fastest switching of a ferromagnet reported to date. Employing an extended microscopic three-temperature model, the temporal dynamics of the exchange coupled ferromagnet–ferrimagnet heterostructure are simulated, qualitatively and quantitatively explaining the experimental switching phenomena. This work experimentally as well as theoretically establishes the mechanism of exchange mediated all-optical switching of ferromagnet-ferrimagnet heterostructures, which can be integrated with a magnetic tunnel junction for efficient reading after ultrafast energy-efficient switching.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Structural Explanation of the Dielectric Enhancement of Barium Titanate Nanoparticles Grown under Hydrothermal Conditions

Abstract When synthesized under certain conditions, barium titanate (BaTiO 3 , BTO) nanoparticles are found to have the non‐thermodynamic cubic structure at room temperature. These particles also have a several‐fold enhanced dielectric constant, sometimes exceeding 6000, and are widely used in thin‐layer capacitors. A hydrothermal approach is used to synthesize BTO nanocrystals, which are characterized by a range of methods, including X‐ray Rietveld refinement and the Williamson–Hall approach, revealing the presence of significant inhomogeneous strain associated with the cubic phase. However, X‐ray pair distribution function measurements clearly show the local structure is lower symmetry than cubic. This apparent inconsistency is resolved by examining 3D Bragg coherent diffraction images of selected nanocrystals, which show the existence of ≈50 nm‐sized domains, which are interpreted as tetragonal twins, and yet cause the average crystalline structure to appear cubic. The ability of these twin boundaries to migrate under the influence of electric fields explains the dielectric anomaly for the nanocrystalline phase.

36 MATERIALS SCIENCE↗