Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “TRAINING PLATFORMS”

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 739 records · Page 41

HydraGNN v5.0

HydraGNN v5.0 expands the code base into a more portable, scalable, and flexible framework for scientific graph learning, with particular strength in atomistic machine-learning interatomic potentials and large-scale distributed training. The release adds Fully Sharded Data Parallel (FSDP) support alongside existing DDP and DeepSpeed paths, including FSDP-aware checkpointing and optimizer integration, and introduces a configurable multi-precision training workflow supporting FP32, BF16, and FP64 across GPUs and Intel XPUs. For atomistic modeling, HydraGNN v5.0 strengthens its MLIP capabilities through dynamic graph construction at every forward pass, energy-conserving force prediction via automatic differentiation, and per-atom energy loss formulations, while extending EGNN models to properly handle periodic boundary conditions. The release also broadens model expressiveness through graph-level attribute conditioning, adds new multi-task and model-parallel extensions such as MACE support and encoder/decoder branch optimization, and expands application coverage with integrated examples for datasets including OC25, Nabla2-DFT, QCML, Open Polymers 2026, and OPF. In parallel, HydraGNN v5.0 improves production readiness through performance optimizations for large-scale runs, stratified sampling and linear-regression preprocessing utilities, and tested installation scripts for DOE supercomputers including Frontier, Aurora, Perlmutter, and Andes. Overall, the release advances HydraGNN as a robust software platform for scalable graph neural networks across materials science, chemistry, and scientific machine learning workflows

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗

Fossil Energy in the Hydrogen Economy – A Carbon-Water-Energy Nexus Adaptive Evaluation Platform

This project develops tools to evaluate fossil energy within the hydrogen economy. In particular, the project seeks to explore the Carbon-Water-Energy Nexus and develop adaptive evaluation tools. The primary goal of this project was to evaluate fossil-based hydrogen production methods with an emphasis on their sustainability, particularly with respect to carbon emissions and water use. The overall technical objectives that guided this evaluation consisted of 1) technology assessment, which sought to establish a comprehensive baseline of existing and emerging technologies for the generation, transportation, storage, and end use of hydrogen derived from fossil fuels, 2) carbon footprint analysis through quantification and analysis of carbon emissions associated with these hydrogen technologies, and identification of novel strategies and technologies that can effectively mitigate the carbon footprint, 3) water intensity assessment by evaluating the water needs (water-energy nexus) associated with different fossil fuel-based hydrogen technologies and proposing strategies to reduce water use, thus ensuring sustainable resource management, and 4) review of regulatory, policy and economic trends were examined to provide an overview of the factors influencing the adoption and integration of fossil fuel-derived hydrogen technologies into current energy markets. Beyond the technical outcomes, this project contributed to 5) education and workforce development through targeted training and educational opportunities for engineering students, preparing the next generation of professionals with technical, economic, and regulatory expertise in the hydrogen economy. Ultimately, this effort was intended to train the future workforce to provide useful insights to policymakers, industry stakeholders, and the broader energy community to facilitate informed decisions and responsible development within the hydrogen economy.

08 HYDROGEN↗

Building problem solving environments with the arches framework

The computational problems that scientists face are rapidly escalating in size and scope. Moreover, the computer systems used to solve these problems are becoming significantly more complex than the familiar, well-understood sequential model on their desktops. While it is possible to re-train scientists to use emerging high-performance computing (HPC) models, it is much more effective to provide them with a higher-level programming environment that has been specialized to their particular domain. By fostering interaction between HPC specialists and the domain scientists, problem-solving environments (PSEs) provide a collaborative environment. A PSE environment allows scientists to focus on expressing their computational problem while the PSE and associated tools support mapping that domain-specific problem to a high-performance computing system. This article describes Arches, an object-oriented framework for building domain-specific PSEs. The framework was designed to support a wide range of problem domains and to be extensible to support very different high-performance computing targets. To demonstrate this flexibility, two PSEs have been developed from the Arches framework to solve problem in two different domains and target very different computing platforms. The Coven PSE supports parallel applications that require large-scale parallelism found in cost-effective Beowulf clusters. In contrast, RCADE targets FPGA-based reconfigurable computing and was originally designed to aid NASA Earth scientists studying satellite instrument data.

99 GENERAL AND MISCELLANEOUS↗

Automatic cataloguing and characterization of Earth science data using SE-trees

In the future, NASA's Earth Observing System (EOS) platforms will produce enormous amounts of remote sensing image data that will be stored in the EOS Data Information System. For the past several years, the Intelligent Data Management group at Goddard's Information Science and Technology Office has been researching techniques for automatically cataloguing and characterizing image data (ADCC) from EOS into a distributed database. At the core of the approach, scientists will be able to retrieve data based upon the contents of the imagery. The ability to automatically classify imagery is key to the success of contents-based search. We report results from experiments applying a novel machine learning framework, based on Set-Enumeration (SE) trees, to the ADCC domain. We experiment with two images: one taken from the Blackhills region in South Dakota; and the other from the Washington DC area. In a classical machine learning experimentation approach, an image's pixels are randomly partitioned into training (i.e. including ground truth or survey data) and testing sets. The prediction model is built using the pixels in the training set, and its performance is estimated using the testing set. With the first Blackhills image, we perform various experiments achieving an accuracy level of 83.2 percent, compared to 72.7 percent using a Back Propagation Neural Network (BPNN) and 65.3 percent using a Gaussain Maximum Likelihood Classifier (GMLC). However, with the Washington DC image, we were only able to achieve 71.4 percent, compared with 67.7 percent reported for the BPNN model and 62.3 percent for the GMLC.

Rymon, Ron↗

Evaluate the impact of sensor accuracy on model performance in data-driven building fault detection and diagnostics using Monte Carlo simulation

The performance of data-driven fault detection and diagnostics (FDD) is heavily dependent on sensors. However, sensor inaccuracy and sensor faults are pervasive in building operation: inaccurate and missing sensor readings deteriorate FDD performance; sensor inaccuracy will also affect the selection of sensor for data-driven FDD in the model training process, which is another key factor of data-driven FDD performance. Sensor accuracy and sensor selection individually are well-studied research topics in this field, but the impact of sensor accuracy on sensor selection and its further impact on FDD performance has not been evaluated and quantified. In this paper, we developed a novel analysis methodology that comprehensively evaluates sensor fault on sensor selection and FDD accuracy. Monte Carlo simulation is applied to deal with multiple stochastic sensor inaccuracy and provide probabilistic analysis results of the impact of sensor inaccuracy on sensor selection and FDD accuracy. This methodology focuses on the net impact of fault states across a full sensor set. The developed methodology can be used for the early-stage sensor design and operation-stage sensor maintenance. Furthermore, a case study is conducted to demonstrate the analysis methodology using a commercial building model crated to Flexible Research Platform located at Oak Ridge National Laboratory, USA.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Critical Experiments at Sandia National Laboratories: Technical Meeting on Low-Power Critical Facilities and Small Reactors

Sandia National Laboratories (SNL) has conducted radiation effects testing for the Department of Energy (DOE) and other contractors supporting the DOE since the 1960's. Over this period, the research reactor facilities at Sandia have had a primary mission to provide appropriate nuclear radiation environments for radiation testing and qualification of electronic components and other devices. The current generation of reactors includes the Annular Core Research Reactor (ACRR), a water-moderated pool-type reactor, fueled by elements constructed from UO2-BeO ceramic fuel pellets, and the Sandia Pulse Reactor III (SPR-III), a bare metal fast burst reactor utilizing a uranium-molybdenum alloy fuel. The SPR-III is currently defueled. The SPR Facility (SPRF) has hosted a series of critical experiments. A purpose-built critical experiment was first operated at the SPRF in the late 1980's. This experiment, called the Space Nuclear Thermal Propulsion Critical Experiment (CX), was designed to explore the reactor physics of a nuclear thermal rocket motor. This experiment was fueled with highly-enriched uranium carbide fuel in annular water-moderated fuel elements. The experiment program was completed and the fuel for the experiment was moved off-site. A second critical experiment, the Burnup Credit Critical Experiment (BUCCX) was operated at Sandia in 2002. The critical assembly for this experiment was based on the assembly used in the CX modified to accommodate low-enriched pin-type fuel in water moderator. This experiment was designed as a platform in which the reactivity effects of specific fission product poisons could be measured. Experiments were carried out on rhodium, an important fission product poison. The fuel and assembly hardware for the BUCCX remains at Sandia and is available for future experimentation. The critical experiment currently in operation at the SPRF is the Seven Percent Critical Experiment (7uPCX). This experiment is designed to provide benchmark reactor physics data to support validation of the reactor physics codes used to design commercial reactor fuel elements in an enrichment range above the current 5% enrichment cap. A first set of critical experiments in the 7uPCX has been completed. More experiments are planned in the 7uPCX series. The critical experiments at Sandia National Laboratories are currently funded by the US Department of Energy Nuclear Criticality Safety Program (NCSP). The NCSP has committed to maintain the critical experiment capability at Sandia and to support the development of a critical experiments training course at the facility. The training course is intended to provide hands-on experiment experience for the training of new and re-training of practicing Nuclear Criticality Safety Engineers. The current plans are for the development of the course to continue through the first part of fiscal year 2011 with the development culminating is the delivery of a prototype of the course in the latter part of the fiscal year. The course will be available in fiscal year 2012.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

A novel transfer learning framework for sorghum biomass prediction using UAV-based remote sensing data and genetic markers

Yield for biofuel crops is measured in terms of biomass, so measurements throughout the growing season are crucial in breeding programs, yet traditionally time- and labor-consuming since they involve destructive sampling. Modern remote sensing platforms, such as unmanned aerial vehicles (UAVs), can carry multiple sensors and collect numerous phenotypic traits with efficient, non-invasive field surveys. However, modeling the complex relationships between the observed phenotypic traits and biomass remains a challenging task, as the ground reference data are very limited for each genotype in the breeding experiment. In this study, a Long Short-Term Memory (LSTM) based Recurrent Neural Network (RNN) model is proposed for sorghum biomass prediction. The architecture is designed to exploit the time series remote sensing and weather data, as well as static genotypic information. As a large number of features have been derived from the remote sensing data, feature importance analysis is conducted to identify and remove redundant features. A strategy to extract representative information from high-dimensional genetic markers is proposed. To enhance generalization and minimize the need for ground reference data, transfer learning strategies are proposed for selecting the most informative training samples from the target domain. Consequently, a pre-trained model can be refined with limited training samples. Field experiments were conducted over a sorghum breeding trial planted in multiple years with more than 600 testcross hybrids. The results show that the proposed LSTM-based RNN model can achieve high accuracies for single year prediction. Further, with the proposed transfer learning strategies, a pre-trained model can be refined with limited training samples from the target domain and predict biomass with an accuracy comparable to that from a trained-from-scratch model for both multiple experiments within a given year and across multiple years.

36 MATERIALS SCIENCE↗

Smartphone-Based Device for Colorimetric Detection of MicroRNA Biomarkers Using Nanoparticle-Based Assay

The detection of microRNAs (miRNAs) is emerging as a clinically important tool for the non-invasive detection of a wide variety of diseases ranging from cancers and cardiovascular illnesses to infectious diseases. Over the years, miRNA detection schemes have become accessible to clinicians, but they still require sophisticated and bulky laboratory equipment and trained personnel to operate. The exceptional computing ability and ease of use of modern smartphones coupled with fieldable optical detection technologies can provide a useful and portable alternative to these laboratory systems. Herein, we present the development of a smartphone-based device called Krometriks, which is capable of simple and rapid colorimetric detection of microRNA (miRNAs) using a nanoparticle-based assay. The device consists of a smartphone, a 3D printed accessory, and a custom-built dedicated mobile app. We illustrate the utility of Krometriks for the detection of an important miRNA disease biomarker, miR-21, using a nanoplasmonics-based assay developed by our group. We show that Krometriks can detect miRNA down to nanomolar concentrations with detection results comparable to a laboratory-based benchtop spectrophotometer. With slight changes to the accessory design, Krometriks can be made compatible with different types of smartphone models and specifications. Thus, the Krometriks device offers a practical colorimetric platform that has the potential to provide accessible and affordable miRNA diagnostics for point-of-care and field applications in low-resource settings.

Krishnan, Tushar (ORCID:0000000342610336)↗

The International Space Station human life sciences experiment implementation process

The selection, definition, and development phases of a Life Sciences flight research experiment has been consistent throughout the past decade. The implementation process, however, has changed significantly within the past two years. This change is driven primarily by the shift from highly integrated, dedicated research missions on platforms with well defined processes to self contained experiments with stand alone operations on platforms which are being concurrently designed. For experiments manifested on the International Space Station (ISS) and/or on short duration missions, the more modular, streamlined, and independent the individual experiment is, the more likely it is to be successfully implemented before the ISS assembly is completed. During the assembly phase of the ISS, science operations are lower in priority than the construction of the station. After the station has been completed, it is expected that more resources will be available to perform research. The complexity of implementing investigations increases with the logistics needed to perform the experiment. Examples of logistics issues include- hardware unique to the experiment; large up and down mass and volume needs; access to crew and hardware during the ascent or descent phases; maintenance of hardware and supplies with a limited shelf life,- baseline data collection schedules with lengthy sessions or sessions close to the launch or landing; onboard stowage availability, particularly cold stowage; and extensive training where highly proficient skills must be maintained. As the ISS processes become better defined, experiment implementation will meet new challenges due to distributed management, on-orbit resource sharing, and adjustments to crew availability pre- and post-increment. c 2001. Elsevier Science Ltd. All rights reserved.

Research↗

A Decision Support System for Extravehicular Operations Under Significant Communication Latency

Humanity hopes to perform extravehicular activities (EVAs) on the surface of Mars; however, several technical and operational challenges must first be overcome. Foremost among these challenges is managing a significant communication latency between Earth and Mars. Current and historical paradigms of EVA operations have required near-real-time communication between the crewmember(s) and Earth-based mission control. Nextgeneration operational paradigms for supporting deep space exploration will necessitate a distributed decision authority system, including delayed Earth-based mission control, the onplanet extravehicular crewmember(s), and intermediate mission support from intravehicular (IV) crewmember(s) within real-time communication range. This latter group is of particular interest: they must provide operations support without the plentiful resources available to mission control on Earth. Thus, NASA is developing the Personalized EVA Informatics and Decision Support (PersEIDS) software platform. PersEIDS is designed to bolster operator situational awareness and offload operator workload by automatically tracking and projecting consumables usage over an EVA timeline, providing real-time probabilistic safety assessments and recommending alternative EVA timeline(s) when the active timeline is not expected to be completed under consumables limits. The PersEIDS concept of operations, use cases, and models will be presented. A limited version of PersEIDS was demonstrated during a three-day-long study where each day a roughly four-hour-long simulated Martian EVA was performed in virtual reality at the NASA Johnson Space Center. The first day was a control trial without PersEIDS support; the second and third days represented different levels of decision support provided by PersEIDS to the IV crewmember acting as mission control. With PersEIDS support, the IV crewmember was able to manage the mission to completion faster and with more remaining consumables; however, additional testing is required to understand confounding factors, e.g., training bias.

Mars↗

PersEIDS: A Biomedical Decision Support System for Extravehicular Operations Under Significant Communication Latency

Humanity hopes to perform extravehicular activities (EVAs) on the surface of Mars; however, several technical and operational challenges must first be overcome. Foremost among these challenges is managing a significant communication latency between Earth and Mars. Current and historical paradigms of EVA operations have required near-real-time communication between the crewmember(s) and Earth-based mission control. Nextgeneration operational paradigms for supporting deep space exploration will necessitate a distributed decision authority system, including delayed Earth-based mission control, the onplanet extravehicular crewmember(s), and intermediate mission support from intravehicular (IV) crewmember(s) within real-time communication range. This latter group is of particular interest: they must provide operations support without the plentiful resources available to mission control on Earth. Thus, NASA is developing the Personalized EVA Informatics and Decision Support (PersEIDS) software platform. PersEIDS is designed to bolster operator situational awareness and offload operator workload by automatically tracking and projecting consumables usage over an EVA timeline, providing real-time probabilistic safety assessments and recommending alternative EVA timeline(s) when the active timeline is not expected to be completed under consumables limits. The PersEIDS concept of operations, use cases, and models will be presented. A limited version of PersEIDS was demonstrated during a three-day-long study where each day a roughly four-hour-long simulated Martian EVA was performed in virtual reality at the NASA Johnson Space Center. The first day was a control trial without PersEIDS support; the second and third days represented different levels of decision support provided by PersEIDS to the IV crewmember acting as mission control. With PersEIDS support, the IV crewmember was able to manage the mission to completion faster and with more remaining consumables; however, additional testing is required to understand confounding factors, e.g., training bias.

Mars↗

Assessing the Needs of NASA's Near Real-Time Earth Observation Products

"The 2017-2027 Decadal Survey for Earth Science and Applications from Space stated that NASA's Earth Science with planned implementation of applications provides sustained earth observations for societal benefits [1]. The Decadal Survey indicated that data latency is invaluable for time-sensitive applications including disaster risk reduction, wildland fire carbon emissions quantification, real-time measurements of the state of the hydrologic systems and many more. Data latency refers to the time between earth observation and data products available to users. During the past 13 years, NASA's Land, Atmosphere Near Real-Time Capability for Earth Observing Systems (LANCE) continues to provide free access to earth observation products that are made available much quicker than routine processing allows. The latency of most LANCE data products is Near Real-time (NRT) which is defined as less than three hours from satellite observations [2]. LANCE is managed by the Earth Science Data and Information System (ESDIS) Project at NASA Goddard Space Flight Center [3], and a User Working Group (UWG) is responsible for providing guidance to LANCE. LANCE data are used by direct users and brokers who add value to the data [4]. NASA Earth Applied Sciences Program (ASP) is one of the primary users of LANCE, which collaborates with partner organizations and provides support to scientists to solve problems in applications of earth observations. ASP promotes the use of LANCE NRT data products to demonstrate applications in decision making, facilitates end-user feedback to the science team to improve data products, and provides information on future demands for research. LANCE supports applications that need a rapid response including detecting wildland fires and volcanic eruptions, tracking smoke, ash and dust plumes, monitoring air quality and tracking extreme weather events such as hurricanes, landslides, and floods. To gather feedback regarding the availability, accessibility and actionability of NASA's NRT data products for societal benefit, three surveys and a few discussions with experts involved in the topic within ASP were conducted from the perspective of users. Feedback has been collected from users who are interested in using low latency NASA data within application communities of agriculture, disasters, water resources, health and air quality, ecological conservation, wildland fires and capacity building. Analysis-ready NRT data products in a variety of formats have been mentioned many times in the collected feedback, especially for applied users with little to no experience using research-grade earth observation products. Users prefer to have products that can be easily integrated into their existing workflows and take their analysis to the data. HDF5 is a commonly used data format for research, but typically requires some conversion to a more friendly format for applications and regular use in decision-making. Users prefer the GeoTIFF data format that can be directly ingested into a GIS mapping software and platform for data analysis and visualization. For example, LANCE’s fire, flood, SO2 and Black Marble Nighttime Blue/Yellow Composite data products have been integrated into NASA Disasters Mapping Portal, which is an GIS-based open data portal, for users in the disaster management community. There are 291 LANCE NRT layers available through GIBS and Worldview, where users can download a snapshot in GeoTIFF format. Operational users expect data to be processed as close to the user as possible. The collected feedback indicates that LANCE fire products within 3 hours latency would meet the needs of the wildland fire community. The ideal latency for volcanic application is 10-15 minutes. Users in Volcanic Ash Advisory Centers (VAAC) reported that the first forecast volcanic product should be issued within 75 minutes from the volcano eruption [5]. Overall, for disaster applications, data latency within 3 hours is useful while latency greater than 12 hours is not timely enough for operational use. Capacity building and training are critical for users to be able to access, interpret and use data products and tools for their decision making, especially for applied users with limited experience using earth observation products. LANCE data products have been used in a number of capacity building projects domestically and internationally [6]. As LANCE continues to bring new products into the system, users request training to utilize LANCE new and upcoming data products and capabilities in their applications. Due to the limitation of bandwidth and downstream flow paths, users in some developing countries need tools to select and download data for a specific area of interest instead of bulk downloads. The collected feedback also shows the lack of available SAR satellite low latency data products. The advantages of SAR to monitor conditions and changes on the ground through darkness, clouds, volcanic ash, and other atmospheric conditions, are appealing to low latency users. For example, terabytes of low latency but cloudy optical images are not helpful in rapidly identifying the extent of flood or fire impacts. LANCE could be complemented with low latency measurements via the upcoming NASA-ISRO Synthetic Aperture Radar (NISAR) mission [7]. Requests for higher spatial resolution products are expressed. A user from the wildland fire management community reported that products with 30-m spatial resolution could be used to detect small fires. The 30-m Landsat OLI fire data is now part of NASA’s Fire Information for Resource Management System (FIRMS) US/Canada [8]. Within the open and free NASA resources, LANCE disseminates NRT data products in a manner that allows them to be accessible and understandable to both scientific and applied users. In many application areas, latency plays an important or even decisive role where low latency earth observations help people to observe areas of interest, detect and track changes in the environment and make timely decisions. NASA’s Earth Applied Sciences Program promotes the use of LANCE NRT products and builds a bridge between application users and research teams. The collected feedback indicates data latency within 3 hours is useful for most of the applications, and shows the needs of user-friendly, analysis-ready products, and requests training on LANCE’s new and upcoming data products. User feedback has been provided to LANCE UWG for guidance and recommendations, and for translating findings into something actionable.

Tian Yao↗

Rancor Integrated Procedure System (RIPS): A Computer-Based Procedure Platform for Advanced Reactor Research

The Rancor Microworld Simulator is a simplified, pressurized water, small modular reactor simulator that includes a multi-unit plant model server, an advanced digital human-machine control interface, and the Rancor Integrated Procedure System (RIPS). Rancor provides a research and development tool that can be used for collecting operator performance data and for prototyping concepts of operations (ConOps) for advanced reactor development. RIPS is meant as a research tool and includes many unique features: (1) RIPS has a robust procedure authoring system. (2) RIPS has the capability to run any of the three IEEE-Std-1786 computer-based procedure types. (3) RIPS can be configured to take on the look and feel of different vendors’ computer-based procedure systems for the purpose of developing and evaluating different ConOps for plant upgrades or new builds. (4) RIPS includes the capability for logging operator procedure use, including integrating procedure logs with Rancor simulator logs, thereby allowing automated data collection of operator scenario runs. (5) RIPS integrates with the Human Unimodel for Nuclear Technology to Enhance Reliability (HUNTER), a dynamic human reliability analysis environment that creates a digital human twin or virtual operator to mimic reactor operator performance. (6) RIPS includes support for automation of plant monitoring and control functions. While RIPS is explicitly built into Rancor, it may also be used with full-scope training simulators. This functionality allows RIPS to be used for existing plants and advanced reactors under development.

99 - GENERAL AND MISCELLANEOUS↗

A Demonstration of Intelligent Container Surveillance Using Stationary and Mobile Camera Platforms

Surveillance of nuclear material storage containers is required to ensure that the container safety boundary is maintained during the service life of the container. As many of these container types are the first and only containment barrier to release protecting the worker, public, and environment, it is imperative to develop robust surveillance tools to characterize container degradation. Machine learning (ML) techniques have matured quickly in the past decade and are slowly becoming a routine application in data analysis. While more popular architectures have been developed with different applications in mind, these can be easily translated to container surveillance requirements. In this study, we developed a ML model based on the Detectron2 framework to identify slip lid containers and common exterior container defects, primarily dents. Commercial-off-the-shelf cameras are combined into a stationary camera array or are mounted to a robotic arm to provide more mobility for positioning needs. Results from the model evaluations on the images indicate that the stationary camera arrays outperform the accuracy of the mobile camera systems for both individual camera and composite image detections. Overall, however, the detection accuracies of the systems fell short of 50% and were less than satisfactory. Future efforts to improve the system involve focusing on a single camera deployable solution that combines a controllable light source that eliminates one of the biggest environmental factors that influence ML model detection performance. Additional training set collection is planned to build a more robust model that enables accurate detection from a wider range of imaging conditions.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Toward Robust Autotuning of Noisy Quantum dot Devices

The current autotuning approaches for quantum dot (QD) devices, while showing some success, lack an assessment of data reliability. This leads to unexpected failures when noisy or otherwise low-quality data is processed by an autonomous system. In this work, we propose a framework for robust autotuning of QD devices that combines a machine learning (ML) state classifier with a data quality control module. The data quality control module acts as a “gatekeeper” system, ensuring that only reliable data are processed by the state classifier. Lower data quality results in either device recalibration or termination. To train both ML systems, we enhance the QD simulation by incorporating synthetic noise typical of QD experiments. We confirm that the inclusion of synthetic noise in the training of the state classifier significantly improves the performance, resulting in an accuracy of 95.0(9)% when tested on experimental data. We then validate the functionality of the data quality control module by showing that the state classifier performance deteriorates with decreasing data quality, as expected. Our results establish a robust and flexible ML framework for autonomous tuning of noisy QD devices.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A frugal CRISPR kit for equitable and accessible education in gene editing and synthetic biology

Equitable and accessible education in life sciences, bioengineering, and synthetic biology is crucial for training the next generation of scientists, fostering transparency in public decision-making, and ensuring biotechnology can benefit a wide-ranging population. As a groundbreaking technology for genome engineering, CRISPR has transformed research and therapeutics. However, hands-on exposure to this technology in educational settings remains limited due to the extensive resources required for CRISPR experiments. Here, we develop CRISPRkit, an affordable kit designed for gene editing and regulation in high school education. CRISPRkit eliminates the need for specialized equipment, prioritizes biosafety, and utilizes cost-effective reagents. By integrating CRISPRi gene regulation, colorful chromoproteins, cell-free transcription-translation systems, smartphone-based quantification, and an in-house automated algorithm (CRISPectra), our kit offers an inexpensive (~$2) and user-friendly approach to performing and analyzing CRISPR experiments, without the need for a traditional laboratory setup. Experiments conducted by high school students in classroom settings highlight the kit’s utility for reliable CRISPRkit experiments. Furthermore, CRISPRkit provides a modular and expandable platform for genome engineering, and we demonstrate its applications for controlling fluorescent proteins and metabolic pathways such as melanin production. We envision CRISPRkit will facilitate biotechnology education for communities of diverse socioeconomic and geographic backgrounds.

59 BASIC BIOLOGICAL SCIENCES↗

Meeting Global Health Needs via Infectious Disease Forecasting: Development of a Reliable Data-Driven Framework

Infectious diseases (IDs) have a significant detrimental impact on global health. Timely and accurate ID forecasting can result in more informed implementation of control measures and prevention policies. To meet the operational decision-making needs of real-world circumstances, we aimed to build a standardized, reliable, and trustworthy ID forecasting pipeline and visualization dashboard that is generalizable across a wide range of modeling techniques, IDs, and global locations. We forecasted 6 diverse, zoonotic diseases (brucellosis, campylobacteriosis, Middle East respiratory syndrome, Q fever, tick-borne encephalitis, and tularemia) across 4 continents and 8 countries. We included a wide range of statistical, machine learning, and deep learning models (n=9) and trained them on a multitude of features (average n=2326) within the One Health landscape, including demography, landscape, climate, and socioeconomic factors. The pipeline and dashboard were created in consideration of crucial operational metrics—prediction accuracy, computational efficiency, spatiotemporal generalizability, uncertainty quantification, and interpretability—which are essential to strategic data-driven decisions. While no single best model was suitable for all disease, region, and country combinations, our ensemble technique selects the best-performing model for each given scenario to achieve the closest prediction. For new or emerging diseases in a region, the ensemble model can predict how the disease may behave in the new region using a pretrained model from a similar region with a history of that disease. The data visualization dashboard provides a clean interface of important analytical metrics, such as ID temporal patterns, forecasts, prediction uncertainties, and model feature importance across all geographic locations and disease combinations. As the need for real-time, operational ID forecasting capabilities increases, this standardized and automated platform for data collection, analysis, and reporting is a major step forward in enabling evidence-based public health decisions and policies for the prevention and mitigation of future ID outbreaks.

60 APPLIED LIFE SCIENCES↗

Z-Target Radiography Postprocessing With A Deep Convolution Neural Network

Analyzing X-ray radiographs is crucial for understanding target behavior in Inertial Confinement Fusion (ICF) and High Energy Density (HED) platforms. However, the density of Magneto Raleigh Taylor (MRT) bands and limitations of target materials often obscure relevant spike growth and density information. To address this issue, machine learning postprocessing techniques can be applied to remove darkened regions in radiography images. In this study, a novel method is presented for removing MRT darkened regions from z-target radiographs using a convolutional neural network (CNN). The CNN, consisting of six layers, treats the darkened regions as noise and employs a mixed loss function and end-to-end frameworks to suppress them while preserving sharpness. The six-layer architecture is designed to effectively learn features when provided with a larger volume of learning space. Each layer is optimized using a mixed loss function that combines a standard loss pixel approach with a multi-scaled structural similarity index loss, which considers luminance, contrast, and structure in local neighborhoods. This approach is particularly beneficial for capturing the stochastic structure of MRT limbs. Due to the limited availability of experimental data, training is conducted using synthetic target radiography from 3D Alegra simulations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗