SEARCH · Search NASA

Results for “data challenges”

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.

SourceResearch connectionFrom yearThrough year

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

At least 433 records · Page 24

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Verifying the Computational Integrity of Power Grid Controls with Zero-Knowledge Proof

The control of future power grids is migrating from a centralized to a distributed/decentralized scheme to enable a massive penetration of distributed energy resources with rising dependence on communication infrastructure. A common assumption made for most existing distributed/decentralized controllers is that local controllers would faithfully follow the designated controller dynamics based on the data received from communication channels. However, with increased probability of cyberattacks on Operational Technology infrastructure, such an assumption could be risky because proper execution of the controller dynamics is then built on trust in secure communication and computation. In this work, we leverage a cryptography technology known as zero-knowledge scalable transparent arguments of knowledge (zk-STARK) to verify the computational integrity of power grid control algorithms, with projected linear dynamics-based control schemes as the initial proof-of-concept test case. The method presented here converts the cybersecurity challenge of data integrity for grid control into a subset of computational integrity.

computational integrity↗

DOE OSTI2021

Performance Portability in the Exascale Computing Project: Exploration Through a Panel Series

Performance portability is a critical issue for the Exascale Computing Project (ECP) because of nontrivial architectural differences between machines available today and those expected at exascale. Many ECP project teams are working toward performance portability, and would expect to benefit from sharing lessons learned, identifying gaps, and discovering opportunities for partnerships. To facilitate this communication, the IDEAS-ECP project partnered with the three focus areas of ECP (application development, software technology, and hardware and integration), and Department of Energy computing facilities, to lead a series of panel discussions on performance portability. The panels were organized around broadly common themes of algorithmic and data locality challenges. In this article, we describe the panel series, its objectives, and perspectives from the various areas of the project. Finally, we also discuss use cases that are distinctive, as well as conclusions drawn from the collective experience of the participants.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Black Swan Event Small-Sample Transfer Learning (BEST-L) and its Case Study on Electrical Power Prediction in COVID-19

The black swan event will usually cause a great impact on the normal operation of society. The scarcity of such events leads to a lack of relevant data and challenges in dealing with related problems. Different situations also make the traditional methods invalid. In this paper, a transfer learning framework and a convolutional neuron network are proposed to deal with the black swan small-sample events (BEST-L). Taking the COVID-19 as a typical black swan event, the BEST-L is utilized to achieve accurate mid-term load forecasting using the relationship between economy and electricity consumption. The experiment results show that the transfer learning model can effectively learn the basic knowledge about the relationship between the adopted input and output data and use a relatively small amount of data during the black swan event to improve the target areas' generalization. The approach and results can provide an effective approach to respond and react to sudden changes quickly and effectively in similar open problems.

black swan event↗

DOE OSTI2023

Testing shell-model interactions at high excitation energy and low spin: Nuclear resonance fluorescence in Ge 74

With their complex low-spin structure, the germanium isotopes have accrued a large number of experimental data which challenge nuclear models. The structure of the 74 Ge isotope is investigated here with nuclear resonance fluorescence as part of an extensive campaign of experimental tests of the shell model in the germanium isotopic chain using a number of complementary techniques. Levels were excited with 3.10 – 5.44 MeV photon beams provided by the High Intensity Gamma-Ray Source at TUNL. Many new levels were identified and their spins, parities, branching ratios, and associated scattering cross section values were determined. Large-scale shell-model calculations, which include jj44b and JUN45 effective interactions developed to describe nuclei in this mass region, predict the new data satisfactorily. Herein this study extends the validation of these two interactions, which are candidates for computation of matrix elements to be used in the interpretation of neutrinoless double-β decay experiments in 76 Ge, into the low-spin, high excitation energy domain.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2022

Topical Group on Application and Industry Community Engagement Frontier Snowmass 2021 (Summary Report)

HEP community leads and operates cutting-edge experiments for the DOE Office of Science which have challenging sensing, data processing, and computing requirements that far surpass typical industrial applications. To make necessary progress in the energy, material, and fundamental sciences, development of novel technologies is often required to enable these advanced detector and accelerator programs. Our capabilities include efficient co-design, which is a prerequisite to enable the deployment of advanced techniques in a scientific setting where development spans from rapid prototyping to robust and reliable production scale. This applies across the design spectrum from the low level fabrication techniques to the high level software development. It underpins the requirement for a holistic approach of innovation that accelerates the cycle of technology development and deployment. The challenges set by the next generation of experiments requires a collaborative approach between academia, industry and national labs. Just a single stakeholder will be unable to deliver the technologies required for the success of the scientific goals. Tools and techniques developed for High Energy Physics (HEP) research can accelerate scientific discovery more broadly across DOE Office of Science and other federal initiatives and also benefit industry applications.

43 PARTICLE ACCELERATORS↗

DOE OSTI2024

Unveiling Temporal Performance Deviation: Leveraging Clustering in Microservices Performance Analysis

As the market for cloud computing continues to grow, an increasing number of users are deploying applications as microservices. The shift introduces unique challenges in identifying and addressing performance issues, particularly within large and complex infrastructures. To address this challenge, we propose a methodology that unveils temporal performance deviations in microservices by clustering containers based on their performance characteristics at different time intervals. Showcasing our methodology on the Alibaba dataset, we found both stable and dynamic performance patterns, providing a valuable tool for enhancing overall performance and reliability in modern application landscapes.

Clustering↗

DOE OSTI2021

Strategic Energy Management Program Persistence and Cost Effectiveness An Analysis of the SEM Program Landscape

This study examines the relationship between strategic energy management (SEM) programs and their persistence and cost effectiveness, with analysis based on interview data from 24 SEM program administrators, SEM program evaluations, and other reports. The 80 interview questions focused on the topics of program design, energy savings, energy savings persistence, cost effectiveness, and customer SEM persistence. The generosity of interview respondents provided a wealth of data, resulting in a report of sufficient length to warrant inclusion of this brief guide of the report structure. The major sections are listed below with brief descriptions. Individual sections of this report are mainly stand-alone and do not require reading of other sections. As a result, there is some duplication between sections, but with differing levels of detail. Executive Summary: Presents three key conclusions of this work with a short description of potential actions to advance the understanding of each key finding. Brief Observations: Lists a large number of bulleted observations resulting from this research, arranged by the five major topic areas included in the interviews. Analysis details are provided in the Analysis of Interview Results section. Foundations for this Research: Provides an overview of SEM, SEM frameworks, SEM programs, the topics of persistence and cost effectiveness, and the focus of this research. Methodology: Details the approach and strategy of this research, providing background information relevant to the formulation of interview questions and the identification of which SEM programs to interview. Observations from Compiled Evaluations and Other Reports: Reports observations from the collection and analysis of program evaluations, annual reports, utility planning documents, and SEM-related white papers. This section, presented in bullet form, highlights challenges in data collection and ultimately a comparison of program practices as they pertain to persistence and cost effectiveness. Analysis of Interview Results: Presents detailed analysis of responses from SEM program administrators, arranged by the five major categories examined: program design, energy savings, energy savings persistence, cost effectiveness, and customer SEM persistence. Interview questions are generally grouped together into subsections when it makes sense to examine them together. SEM Programs Challenge Traditional Cost-Effectiveness Metrics: Details an analysis based on five key factors showing that applying traditional cost-effectiveness metrics to SEM programs is not straightforward. This invites the opportunity to consider whether traditional cost-effectiveness metrics are applicable to SEM programs, either individually or at large. Resolution of Research Hypotheses: Tabulates a set of hypotheses that were developed to address the fundamental nature of the research at hand. Analysis of responses to multiple questions informs an understanding of each hypothesis and can be used to better understand the SEM program environment at large.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2025

Predicting the von Neumann entanglement entropy using a graph neural network

Calculating the von Neumann entanglement entropy from experimental data is challenging due to its dependence on the complete wavefunction, forcing reliance on approximations such as classical mutual information (MI). We propose a machine learning approach using a graph neural network to predict the von Neumann entropy directly from experimentally accessible bitstrings. We test this approach on a Rydberg ladder system and achieve a mean absolute error of $3.6\,\times 10^{-3}$ when evaluating within the training range on a dataset with entropy values ranging from 0 to 1.9. The model achieves a mean absolute percentage error of 1.44% and outperforms MI-based bounds. When tested beyond the training range, the model maintains reasonable accuracy. Furthermore, we demonstrate that fine-tuning the model with small datasets significantly improves performance on data outside the original training range.

graph neural networks↗

DOE OSTI2025

Unsupervised anomaly clustering via offset alignment in multivariate grid sensing data

Modern industries increasingly rely on multi-sensor technologies to acquire complex, high-dimensional data streams, enabling advanced monitoring and control systems. One critical application is online anomaly detection in electrical smart grids, where multivariate and multimodal sensing technologies play a vital role. However, detecting anomalies in such time-series data is challenging due to their inherent temporal dependencies and stochastic behavior. Traditional approaches based on supervised and semi-supervised learning methods depend on labeled datasets, which are often unavailable in real-world scenarios. While unsupervised methods have emerged as promising alternatives, these methods are highly susceptible to noise and outliers commonly present in sensing applications. Furthermore, deep learning-based anomaly detection methods, despite their performance, are often criticized for their black-box nature, limiting their applicability in safety-critical and online environments where interpretability and explainability are paramount. In this work, we propose an unsupervised anomaly clustering method leveraging a cyclic alignment-based offset detection algorithm for multivariate time-series signals. The proposed method is applied to multivariate data collected from vibrational, voltage, and magnetic field sensors deployed in a local grid substation. Our results demonstrate the robustness of the algorithm in accurately clustering various anomalies/events across different sensing modalities. Additionally, we compare the effectiveness of the proposed approach against a simple pattern-based anomaly detection method, which performs well for univariate data but fails to generalize to multivariate and multimodal time-series data.

Mukherjee, Subrata [ORNL] (ORCID:0000000309930338)↗

NASA NTRSDate not supplied

The INSTEP Monitoring Network: Merging High-and-Low Cost Measurements to Characterize California Wildfires

Despite challenges with data quality and scope, low-cost sensor networks have skyrocketed in popularity over the last 15 years, making air quality data available on refined spatial scales. More recently, studies have leveraged both high and low-quality instruments to create stronger “hybrid” models, with most studies focusing on particulate matter. Low-cost measurements typically represent ground-level emissions only, providing context for human health issues from climate change-driven events such as wildfires. Since low-cost sensors’ capabilities are localized, daily events and microclimates tend to dominate the data rather than larger regional or atmospheric trends. Likewise, their low cost explains their high uncertainty. In contrast, some regulatory-grade instruments produce column measurements as well, providing reliable information on a broader scope. To bridge this gap while expanding into gas-phase measurements, we deployed 12 air quality sensor packages in California, USA during the 2022 wildfire season. These INSTEP (Inexpensive Network Sensor Technology Exploring Pollution) monitors measure carbon monoxide (CO), carbon dioxide (CO2), ozone (O3), nitrogen dioxide (NO2), and several hydrocarbons including methane (CH4) and formaldehyde (HCHO). Half of the monitors were co-located with remote sensing spectrometers: NASA Pandora and Total Column Carbon Observing Network (TCCON). The overlap in pollutants includes NO2, O3, and HCHO between the INSTEP monitors and the Pandora column measurements. TCCON covers column CO, CO2, and CH4, rounding out our comparison. Most of the monitors were distributed throughout the San Francisco Bay area, and an additional three were located within 100 km of Los Angeles. The sites ranged in geographic and population characteristics, including desert, mountainous, coastal, and urban locations. Since varying environmental conditions such as temperature and pressure are known to challenge sensor performance, we will apply newer sensor “calibration” techniques meant to combat this. We will normalize our sensor signals by z-scoring them prior to applying a single calibration model in the form of multivariate linear regression or an artificial neural network. While this technique has been validated for the hydrocarbon and ozone sensor types (metal oxide), it has not yet been tested on electrochemical and non-dispersive infrared sensors, which are also used in the INSTEP monitors. This will serve as a test to see if this normalization technique – or another – is most effective in accounting for environmental differences among sensors. Related data analysis efforts have found success with a variety of geospatial analysis techniques, including weighted network models in which high-quality instruments are given higher weights than their low-cost counterparts. Our preliminary analysis will focus on kriging, which uses a Gaussian algorithm to assign weights, providing estimated pollution levels at locations between monitors. Smoke trajectory and evolution will also be considered using both measurement types. We also aim to baseline subtract our emission estimates from each region to determine which portion of emissions are regional and local, further characterizing burn differences in northern and southern California fires. Future directions include using INSTEP jointly with TEMPO satellite data, and mobile deployments on aircraft and uncrewed aerial vehicles (UAV).

Low-cost sensors↗

NASA NTRS2016

Characterization of Operational Performance of Ka-Band Links in Deep Space Network

This paper presents a characterization of performance of 32-GHz Ka-band link in an operational environment. The data come from tracking of Kepler spacecraft by the NASA Deep Space Network (DSN) in the past few years. Ka-band link offers a significant signal-to-noise advantage compared to the more common X- or S-band; however, it is subject to more signal fluctuation caused by the weather. Kepler is the first mission supported by the DSN that solely rely on the higher deep space Ka-band communications link to return its high-rate science data. A well characterization of the operational performance of Kepler would benefit future Ka-band missions, especially for those operating with smaller link margin. The study examines how weather conditions at the DSN facilities (e.g., winds, clouds, rains) affect the received signal, particularly on telemetry data. It addresses questions such as how often the weather affects the link and how much degradation the link could suffer. Among the 22 Ka-band passes in 2012, heavy clouds affected one pass and two passes were impacted by high winds. The adverse weather caused at time as much as 3-dB instantaneous change in the signal to noise ratio (SNR). Estimates of degradation to the SNR as a function of wind speeds are captured based on observations. This study also quantifies the probability distribution of the variation of received signal power. With such information, future missions can better plan their link design. An optimal link design aims for just having enough margin to realize the data return with a targeted probability, with neither having too much margin that it reduces the downlink data rate nor insufficient reserve that causes frequent data outages. The study tries to provide some answers to the questions such as how much does SNR vary from one tracking pass to the next, and what is the cumulative distribution of signal fluctuation. Regarding the signal variation over many tracking passes, we found that the averaged symbol SNR (SSNR) for each pass changed by as much as 4 dB. Some variations were due to geometry such as the changing distance between spacecraft and Earth and different antenna pointing elevations. Other variations seem to be random in nature, reflecting the randomness of operational environments. Some passes were found to be quite stable, with a standard deviation of symbol SNR around 0.25 dB while others had greater variation, up to 1.4 dB. Overall, the cumulative distribution of the symbol SNR reflects that 50% of the fluctuations was less than 0.6 dB, 90% of fluctuation was within 1.6 dB and 95% within 2.2 dB. Some of the challenges in data processing, validation and modeling were captured in this paper. We observed some inconsistent measurements where operational data significantly deviate from a normal expectation. These inconsistent measurements made it hard to develop an accurate and consistent performance model, such as the degradation of signal SNR as a function of high wind. The difficulty was compounded by the fact that there were only a few Ka-band passes observed in high winds.

Pham, Timothy↗

DOE OSTI2025

Leveraging differentiable programming in the inverse problem of neutron stars

Neutron stars (NSs) probe the high-density regime of the nuclear equation of state (EOS). However, inferring the EOS from observations of NSs is a computationally challenging task. Here, in this work, we efficiently solve this inverse problem by leveraging differential programming in two ways. First, we enable full Bayesian inference in under one hour of wall time on a GPU by using gradient-based samplers, without requiring pretrained machine learning emulators. Moreover, we demonstrate efficient scaling to high-dimensional parameter spaces. Second, we introduce a novel gradient-based optimization scheme that recovers the EOS of a given NS mass-radius curve. We demonstrate how our framework can reveal consistencies or tensions between nuclear physics and astrophysics. First, we show how the breakdown density of a metamodel description of the EOS can be determined from NS observations. Second, we demonstrate how degeneracies in EOS modeling using nuclear empirical parameters can influence the inverse problem during gradient-based optimization. Looking ahead, our approach opens up new theoretical studies of the relation between NS properties and the EOS, while effectively tackling the data analysis challenges brought by future detectors.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2022

A laser powder bed fusion system for operando synchrotron x-ray imaging and correlative diagnostic experiments at the Stanford Synchrotron Radiation Lightsource

Laser powder bed fusion (LPBF) is a highly dynamic multi-physics process used for the additive manufacturing (AM) of metal components. Improving process understanding and validating predictive computational models require high-fidelity diagnostics capable of capturing data in challenging environments. Synchrotron x-ray techniques play a vital role in the validation process as they are the only in situ diagnostic capable of imaging sub-surface melt pool dynamics and microstructure evolution during LPBF-AM. In this article, a laboratory scale system designed to mimic LPBF process conditions while operating at a synchrotron facility is described. The system is implemented with process accurate atmospheric conditions, including an air knife for active vapor plume removal. Significantly, the chamber also incorporates a diagnostic sensor suite that monitors emitted optical, acoustic, and electronic signals during laser processing with coincident x-ray imaging. The addition of the sensor suite enables validation of these industrially compatible single point sensors by detecting pore formation and spatter events and directly correlating the events with changes in the detected signal. Experiments in the Ti–6Al–4V alloy performed at the Stanford Synchrotron Radiation Lightsource using the system are detailed with sufficient sampling rates to probe melt pool dynamics. X-ray imaging captures melt pool dynamics at frame rates of 20 kHz with a 2 µm pixel resolution, and the coincident diagnostic sensor data are recorded at 470 kHz. Furthermore, this work shows that the current system enables the in situ detection of defects during the LPBF process and permits direct correlation of diagnostic signatures at the exact time of defect formation.

36 MATERIALS SCIENCE↗

NASA NTRSDate not supplied

Unveiling Urban Pollution Patterns from a Bird’s-Eye View

Low-earth orbiting satellites have been observing the global distribution of nitrogen dioxide (NO2) since the 1990s and these observations have been used to identify pollution source regions, temporal trends, and links to health effects and community disparities. However, historically, space-based NO2 observations have been coarse in spatial and temporal resolution resulting in challenges in data interpretation and product validation; therefore, NASA developed airborne capabilities and measurements strategies to observe ultraviolet-visible (UV-VIS) trace gases at high temporal and spatial resolutions to unveiling never before seen spatiotemporal patterns over major urban hubs during recent air quality field studies. In this talk, Dr. Judd will share the story of these field studies with datasets collected from NASA’s airborne spectrometers, GeoTASO and GCAS, and how they integrate in with other space- and ground-based instrumentation to illustrate the conceptual model of air quality episodes in coastal urban environments, contribute to satellite validation and regional chemical transport model evaluation. These imagers have the capability to observe UV-VIS absorbing species, typically NO2 and HCHO, at spatial resolutions as fine as 250 m when observing from the upper-troposphere. Flight strategies allowed for the collection of repeated raster samples over urban areas multiple times per flight day in cities like Chicago, Los Angeles, and New York City, capturing never before seen pictures of the diurnal evolution of emissions, chemistry, and meteorology associated with NO2 heterogeneity. These airborne efforts help local air quality management better understand their air quality challenges as well as benefit the larger air quality community in preparation for geostationary air quality observations and the development of the global air quality observing system.

Laura Judd↗

DOE OSTI2025

Circularity Futures Workshop Series: Summary Report

The aim of this report is to synthesize key feedback received from the three-part Circularity Futures workshop series held in Spring 2024. The workshop series was conducted by the National Renewable Energy Laboratory (NREL) on behalf of U.S. Department of Energy, Office Energy Efficiency and Renewable Energy (EERE), and was broken into three workshops: Workshop 1 - Circularity Analysis Needs and Priorities; Workshop 2 - Circularity Metrics and Indicators; and Workshop 3 - Circularity Data. Together, the workshops focused on identifying the existing priorities and gaps in the circularity modeling space, understanding different stakeholders' use and interpretation of circularity metrics and indicators, identifying common data gaps and data quality challenges, and assessing the robustness of available solutions. The workshop series brought a diverse group of stakeholders - including representatives from U.S. government offices, national labs, nonprofit organizations, industry, and academia - to collect first-hand feedback on needs, priorities, challenges and opportunities in the circularity modeling and analysis space. The workshop discussions highlighted numerous common needs, priorities and challenges among the interviewed groups. Several topics were frequently discussed, including: 1) Circularity as a pathway for sustainable economic growth: While circularity is generally defined in terms of resource conservation and reducing wasteful disposal of materials, participants agreed that circular strategies should serve broader economic, environmental, and social goals. It is therefore crucial for circularity analysis to look beyond waste reduction and instead evaluate a variety of impact metrics such as cost savings, job creation, air quality, and pollutant emissions. Mutli-criteria decision-making frameworks may be useful for making sense of disparate metrics and evaluating tradeoffs between impact categories.; 2) Economic and social factors are not well understood: Underdevelopment of existing end-of-life (EOL) management infrastructure, inconsistent standardization codes and policy space in reusing recycled content, and suboptimal collection and sorting strategies collectively contribute to uncertainty about the economic potential of circular pathways. The latter observation is consistent among all technologies but more emphasized for renewable energy systems. Social impacts of circularity practices are less understood and less researched than other sustainability aspects.; 3) Inconsistent methods for assessing emerging technologies: LCA and TEA results vary widely depending on the assumptions made with regards to market adoption of new technologies. Emerging technologies suffer limited availability of data needed to conduct a robust circularity analysis. Yet, understanding projected impacts of proposed nascent technology is a key need for different stakeholder groups.; and 4) Lack of temporally and geospatially explicit data: There is a need for open data that represents variations in circularity technologies over time and location. The lack thereof leads to aggregated and potentially misrepresented results in circularity analysis. Sensitivity analyses should be included to verify whether options perceived as more sustainable align with real-world practices.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

NASA NTRS1995

Geostatistical Approaches for Spatial Estimation of Vegetation Quantities Using Ground and Image Data

A major challenge in the study of the earth system is the mapping of vegetation quantities over large regions. Geostatistical methods, such as cokriging and stochastic simulation, have the potential to exploit more fully both remotely sensed data and ground information and improve the spatial estimation of vegetation variables over traditional regression methods. A synthetic example constructed from imaging spectrometer data allows a useful comparison among regression, cokriging and a simple probability-field method. A range of linear relationships between direct (sampled) and ancillary images is used. The lowest root-mean-square-error is achieved with cokriging until the correlation between direct and ancillary data exceeds .89, at which point regression is the superior estimator. Probability-field simulation gives a range of possible realizations, most more precise than those from regression. The relationship between ground measurements and image data is thus demonstrated to be one of the critical factors in the choice of a spatial estimation method.

Dungan, Jennifer L.↗

NASA NTRS2015

3-D CFD Simulation and Validation of Oxygen-Rich Hydrocarbon Combustion in a Gas-Centered Swirl Coaxial Injector using a Flamelet-Based Approach

Injector design is a critical part of the development of a rocket Thrust Chamber Assembly (TCA). Proper detailed injector design can maximize propulsion efficiency while minimizing the potential for failures in the combustion chamber. Traditional design and analysis methods for hydrocarbon-fuel injector elements are based heavily on empirical data and models developed from heritage hardware tests. Using this limited set of data produces challenges when trying to design a new propulsion system where the operating conditions may greatly differ from heritage applications. Time-accurate, Three-Dimensional (3-D) Computational Fluid Dynamics (CFD) modeling of combusting flows inside of injectors has long been a goal of the fluid analysis group at Marshall Space Flight Center (MSFC) and the larger CFD modeling community. CFD simulation can provide insight into the design and function of an injector that cannot be obtained easily through testing or empirical comparisons to existing hardware. However, the traditional finite-rate chemistry modeling approach utilized to simulate combusting flows for complex fuels, such as Rocket Propellant-2 (RP-2), is prohibitively expensive and time consuming even with a large amount of computational resources. MSFC has been working, in partnership with Streamline Numerics, Inc., to develop a computationally efficient, flamelet-based approach for modeling complex combusting flow applications. In this work, a flamelet modeling approach is used to simulate time-accurate, 3-D, combusting flow inside a single Gas Centered Swirl Coaxial (GCSC) injector using the flow solver, Loci-STREAM. CFD simulations were performed for several different injector geometries. Results of the CFD analysis helped guide the design of the injector from an initial concept to a tested prototype. The results of the CFD analysis are compared to data gathered from several hot-fire, single element injector tests performed in the Air Force Research Lab EC-1 test facility located at Edwards Air Force Base.

Richardson, Brian↗

DOE OSTI2025

PySIDT: Subgraph Isomorphic Decision Trees for Molecular Property Prediction

Accurate molecular property prediction is important across all fields of chemistry. Deep neural networks (DNNs) have become increasingly popular due to their ability to train automatically, avoiding the incredibly tedious process of constructing and extending traditional property estimation schemes. However, DNNs require large amounts of training data, are challenging to interpret, require large amounts of memory to load even during inference, and have severe difficulties incorporating qualitative chemical knowledge, which are often desired for molecular property prediction tasks. Here, in this study, we present PySIDT (https://github.com/zadorlab/PySIDT), a software for training and running inference on Subgraph Isomorphic Decision Trees (SIDTs). SIDTs are graph-based decision trees made of nodes associated with molecular substructures. Inference is done by descending target molecular structures down the decision tree to nodes with matching subgraph isomorphic substructures and making predictions based on the final (most specific) nodes matched. SIDTs scale down well to dataset sizes much smaller than is feasible for DNNs. As trees of molecular substructures, SIDTs are inherently readable and easy to visualize, making them easy to analyze. They are also straightforward to extend and retrain, facilitate uncertainty estimation, and enable easy integration of expert knowledge. We demonstrate the SIDT approach discussing its application to a diverse range of molecular prediction tasks: rate coefficient estimation, diffusion coefficient estimation, thermochemistry estimation, transition state bond stretch prediction, p K a prediction, stability of molecular structures, stability of surface structures, and prediction of surface lateral interaction energetics. Additionally, we demonstrate the power of the SIDT algorithms in two direct learning curve vanilla comparisons with the popular DNN-based software Chemprop and the popular gradient boosted trees-based software XGBoost on enthalpy of formation and rate coefficient prediction tasks. In particular, in the enthalpy of formation case, vanilla PySIDT is able to outperform vanilla Chemprop and XGBoost across the full range of training/validation set sizes out to 11,560 data points.

Johnson, Matthew Sean [Sandia National Laboratorie↗