Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Intelligent Adaptive Systems”

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 217 records · Page 12

Reinforcement Learning‐Based Adaptation of Grid Following Inverter's Internal Controller to Networked Microgrids' Strengths

The varying topological configurations, generator commitments and dispatches, and dynamic load demand lead to changing system's strengths during the operations of networked microgrids. When the system's strengths significantly change, the fixed control gains at large devices may result in unsatisfactory system performance; this necessitates the tuning of the control gains at large devices to adapt to the changing system's strengths. In this paper, observer-based reinforcement learning (RL) is utilised to automatically tune the proportional-integral (PI) gains of phase lock loop (PLL) controller of grid-following (GFL) inverters to adapt to the changing strengths of microgrids and networked microgrids. The RL agent in this framework augments an observer predicting system's strengths, from which the RL control policy will adjust accordingly to tune the PLL controller's gains towards the system's strengths. Also, to enhance the control performance, the recently introduced Barrier function-based RL framework is leveraged for the design of reward function to prevent the high frequency nadir. An operational 26 kV electric distribution system, which is modelled as networked microgrids, is used to illustrate the need and effectiveness of the proposed RL-tuned control.

frequency response↗

Latest Developments in PM Systems and Methods and their Application to WM Projects - 20364

How do you improve on an industry that is well established yet often misses the mark on key metrics demonstrating effective planning and execution of project scope, schedule and budget? With the increased complexity of DOE projects relative to either poorly kept records of what wastes are stored or the lack of specific knowledge relative to the waste concentrations/quantities, this is a question continually asked and investigated by many. In fact, in 2017, the GAO's report to Congress indicated that although the DOE has made some progress on its monitoring effectiveness and demonstrating progress criteria, these goals were not met and there needs to be more improvement and monitoring. This paper will offer some insights into the latest developments in project management methodologies and the possible application of these to waste management projects. Traditional data generated via a project schedule that utilizes critical path methodology (CPM) and additional earned value management (EVM) data needs to be accurate and enable the project manager to better understand the probability of project success. Although this data has readily been available on most large-scale projects, success as measured by various stakeholder groups including the client, project team, regulators, etc. is less than optimal. Therefore, an urgent need exists to explore how traditional project data (CPM and EVM) coupled with new methodologies allows for better planning and execution by the project team. Many new approaches/methodologies can even be used to predict the success of options prior to full implementation on the actual project. Three approaches/methodologies will be discussed: - Artificial Intelligence; - Change Management; - Blend of Traditional Management (Gantt chart) and Agile Methodologies. As market forces change, so will the direction of project management. The days of 'one size fits all' are no longer valid. A combination of predictive and adaptive approaches in the management of projects is essential to enable execution of project deliverables on time, under budget, and within the contractual scope. This paper is intended to provide the reader with insight to the close relationship of project management trends and business management trends, and to provoke critical thinking about what it means to align project management strategy with an environment of non-stop innovation. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Navigating Uncertainty: Challenges in Visualizing Ensemble Data and Surrogate Models for Decision Systems

Uncertainty visualization plays a critical role in transforming ensemble simulation data into actionable insights by effectively communicating various dimensions of uncertainty within a system. The emergence of artificial intelligence-driven surrogate models trained on multirun ensemble data offers a transformative opportunity to replace computationally intensive simulations with fast estimates, enabling users to explore data spaces with unprecedented depth and interactivity. However, integrating ensemble data and surrogate models into decision-making workflows and tools introduces novel challenges for uncertainty visualization. These include reconciling and clearly communicating the unique uncertainties associated with ensembles and their surrogate model estimates, and leveraging these approximations to inform actionable decisions. This work explores these challenges in the context of high-dimensional data visualization, bridging discrete datasets with their continuous representations and addressing the complexities of systems that support iterative navigation between input and output spaces. We evaluate the role of uncertainty visualization in fostering intuitive, actionable interactions and identify critical hurdles in advancing this frontier of computational simulation.

97 MATHEMATICS AND COMPUTING↗

Real Time Predictive and Adaptive Hybrid Powertrain Control Development via Neuroevolution

The real-time application of powertrain-based predictive energy management (PrEM) brings the prospect of additional energy savings for hybrid powertrains. Torque split optimal control methodologies have been a focus in the automotive industry and academia for many years. Their real-time application in modern vehicles is, however, still lagging behind. While conventional exact and non-exact optimal control techniques such as Dynamic Programming and Model Predictive Control have been demonstrated, they suffer from the curse of dimensionality and quickly display limitations with high system complexity and highly stochastic environment operation. This paper demonstrates that Neuroevolution associated drive cycle classification algorithms can infer optimal control strategies for any system complexity and environment, hence streamlining and speeding up the control development process. Neuroevolution also circumvents the integration of low fidelity online plant models, further avoiding prohibitive embedded computing requirements and fidelity loss. This brings the prospect of optimal control to complex multi-physics system applications. The methodology presented here covers the development of the drive cycles used to train and validate the neurocontrollers and classifiers, as well as the application of the Neuroevolution process.

33 ADVANCED PROPULSION SYSTEMS↗

Extreme Heterogeneity 2018 - Productive Computational Science in the Era of Extreme Heterogeneity: Report for DOE ASCR Workshop on Extreme Heterogeneity

The 2018 Basic Research Needs Workshop on Extreme Heterogeneity identified five Priority Research Directions for realizing the capabilities needed to address the challenges posed in this era of rapid technological change. This report captures the outcomes of that workshop. In the context of extreme heterogeneity, it defines basic research needs and opportunities in computer science research to develop smart and trainable operating and runtime systems, programming environments, and predictive tools that will make future systems easier to adapt to scientists’ computing needs and easier for facilities to deploy securely.

97 MATHEMATICS AND COMPUTING↗

Autonomous Aerial Power Plant Inspection in GPS-denied Environments

Inspection of coal-fired power plants is frequently dangerous, includes difficult places to reach, and can turn expensive due to the downtimes and cost of inspection crew. Robotic systems have shown capabilities to address some of these issues, but most of the current robotic inspection technology in power plants is designed for specific components. Conversely, recent advances in machine vision have empowered aerial platforms for long-range, remotely-controlled, GPS-based inspections of industrial plants. This capability has led to wide spread utilization of aerial robots (commonly termed Drones, UVS or UAS) platforms for inspection in less challenging environments where both collision avoidance, and GPS reception are not significant issues. The challenge in adapting airborne technology for power plant inspection lies in internal structures and the complex network of piping, and distribution systems, which impose significant risks for collision and can hinder the reception and transmission of GPS signals. The current state of the art in aerial inspection technology within the energy sector is controlled via radio control, and utilizes GPS-based navigation, for inspection of large-scale plants such as offshore platforms and wind turbine parks. Nevertheless, close-range and autonomous inspection in the GPS-denied environments of power plants has not yet been achieved, as it requires precise guidance and navigation with real-time situational awareness and obstacle avoidance capabilities. This endeavor introduced the use of rotary wing flying robots, due to their station keeping and vertical take-off capabilities for power plant components inspection. To enable close quarter inspection two methods were used. One method uses the 3D CAD (Three-dimensional Computer-Aided Design) model of the asset to inspect to generate the UAV’s inspection path. To acquire, analyze and process the 3D model, first, the STL file is produced to obtain surface points and vectors normal to the surface. Later, by introducing other variables such as wall offset and a controlled trajectory between each outline and each subsequent layer, the flight path is generated. The proposed framework will generate a path that will pass as close as desired from the surface and navigate in intricate environments. A second method, use advanced manufacturing techniques such as CNC (Computer Numerical Control) and additive manufacturing. Once the inspection flight path is obtained, vision-based navigation systems are employed to have the UAV autonomously tracking the provided trajectory. Finally, Artificial Intelligence-enabled developments are in charge of detecting cracks and corrosion in structural components of power plants. The proposed methods are validated in simulations, laboratory and industrial setups, where it is shown that the developed systems acting together enable close-quarter autonomous aerial inspection and mapping in power plant assets. The system can be further improved by adding more sensors to navigate in different GPS-denied environments, with non-homogeneous lighting conditions, dust and in general situations where vision-based systems may fail.

01 COAL, LIGNITE, AND PEAT↗

An Intelligent Adaptable Monitoring Package. Final Report

The “Intelligent Adaptable Monitoring Package” project was a four-year effort that demonstrated the feasibility of integrated sensing packages at tidal and wave energy sites. Such integration is generally required by the breadth of sensors required to understand environmental effects at marine energy sites and the operational difficulty of deploying, maintaining, and recovering such sensors. Over the course of the project, the Adaptable Monitoring Package (AMP) was deployed in multiple settings, each corresponding to a project budget period: - Budget Period 1: Demonstration of cabled deployment at Pacific Northwest National Laboratory’s Marine Science Laboratory. The deployment highlighted AMP hardware endurance over a 4-month deployment in a tidally-dominated environment and laid the groundwork for machine learning algorithms to detect and classify targets present in active sonar data. - Budget Period 2: Demonstration of an autonomous deployment at PacWave South off the coast of Newport, Oregon. The deployment highlighted the stability of AMP hardware and software, with the autonomous package collecting data on a duty cycle over a 1.5-month deployment. - Budget Period 3: Demonstration of an autonomous deployment powered by a wave energy converter at the U.S. Navy’s Wave Energy Test Site. The deployment highlighted the potential of wave energy to power ocean observatories and led to the development of machine learning algorithms to detect and classify targets in optical camera data. In aggregate, this project’s greatest success was demonstrating the AMP’s flexibility in a range of deployment scenarios. Each budget period represented a “first of a kind” demonstration of integrated instrumentation – cabled AMP, autonomous AMP, wave-energy powered AMP – and each deployment helped to identify and set goals for the next. Further, despite the exploratory nature of these deployments, each one achieved high system up-time and proved that flexible integration of multiple sensors in a single package represents a viable strategy for marine energy environmental monitoring. The key lessons learned from the project are: - Without continuous power, either from a shore cable or in situ source, many of the benefits of integration are lost (If continuous power is not available, the ability to detect rare events is lost, as is the ability to minimize the risk of behavioral changes through adaptive sensing. However, even on a duty cycle, there is still value in being able to acquire synchronous data from multiple sensors.); and - Observations from a moving platform present substantially greater data processing challenges than those from stationary platforms. Finally, these deployments also demonstrate an important truth: successful integration alone does not guarantee that relevant data are collected. To grow the knowledge base about environmental interactions with marine energy converters, integrated systems, like the AMP, need to include the right sensor mix and connect the data pipelines to effective processing algorithms. These deployments establish a strong foundation for future collaborations with the environmental research community: not only to understand the environmental effects of marine energy, but also to improve our general ability to study life in the sea.

16 TIDAL AND WAVE POWER↗

The EGS Collab Project – Stimulations at Two Depths

The EGS Collab project, supported by the US Department of Energy, is performing intensively monitored rock stimulation and flow tests at the 10-m scale in an underground research laboratory to address challenges in implementing enhanced geothermal systems (EGS). Data and observations from the field tests are compared to simulations to understand processes and build confidence in numerical modeling of the processes. We have completed Experiment 1 (of 3), which examined hydraulic fracturing in a well-characterized underground fractured phyllite test bed at a depth of approximately 1.5 km at the Sanford Underground Research Facility (SURF) in Lead, South Dakota. Testbed characterization included fracture mapping, borehole acoustic and optical televiewers, full waveform sonic, conductivity, resistivity, temperature, campaign p- and s-wave investigations and electrical resistance tomography. Borehole geophysical techniques including passive seismic, continuous active source seismic monitoring, electrical resistance tomography, fiber-based distributed strain, distributed temperature, and distributed acoustic monitoring, were used to carefully monitor stimulation events and flow tests. More than a dozen stimulations and nearly one year of flow tests were performed. Quality data and detailed observations were collected and analyzed during stimulation and water flow tests using ambient temperature and chilled water. We achieved adaptive control of the tests using real-time monitoring and rapid dissemination of data and near-real-time simulation. More detailed numerical simulation was performed to answer key experimental design questions, forecast fracture propagation trajectories and extents, and analyze and evaluate results. Data are freely available from the Geothermal Data Repository. Experiment 2 examines the potential for hydraulic shearing in amphibolite at a depth of about 1.25 km at SURF. This site has a different set of stress and fracture conditions than Experiment 1. The Experiment 2 testbed consists of nine subhorizontal boreholes configured in two fans of two boreholes which surround the testbed and contain grouted-in electrical resistance tomography, seismic sensors, active seismic sources and distributed fiber sensors. A “five-spot” set of test wells that extends from a custom mined alcove includes an injection well and four production/monitoring wells. The testbed was characterized geophysically and hydrologically, and three stimulations have been performed using the Step-Rate Injection Method for Fracture In-Situ Properties (SIMFIP) tool to measure strains, and a new strain quantifying tool (downhole robotic strain analysis tool -DORSA) was deployed in a monitoring hole during stimulation. Real-time data were broadcast during stimulations to allow real-time response to arising issues.

EGS Collab, Enhanced Geothermal Systems, EGS, fiel↗

Explainable Artificial Intelligence Technology for Predictive Maintenance

The domestic nuclear power plant fleet has relied on labor-intensive and time-consuming preventive maintenance programs, thus driving up operation and maintenance costs to achieve high-capacity factors. Artificial intelligence and machine learning can help simplify complex problems, such as diagnosing equipment degradation, to enable more effective decision-making. Benefits will be felt not only within existing analog and digital instrumentation and control, but also work processes, the integration of people with technology, and most importantly, the business case. Together, these hold promise to make nuclear power more efficient and reduce costs associated with operation and maintenance. While the artificial intelligence and machine learning technologies hold significant promise in the nuclear industry, there are challenges or barriers to their adoption. This report outlines the those different machine learning adoption barriers (categorized as historical, technical, economic, regulatory, and user) that the industry must overcome to realize the full benefits of artificial intelligence and machine learning capabilities for long-term economic sustainability. This report also provides solutions for some of these barriers by focusing on improving the explainability of machine learning to encourage trust from the end-user. Trust and explainability are essential to machine learning adoption. This report focuses on research-developed solutions to some of these barriers while analyzing a non-safety-related system, namely the circulating water system. This system frequently experiences waterbox fouling which our models preemptively diagnoses then explains to the operator how those conclusions were reached. This report presents and discusses the inherent trade-off between machine learning performance (in terms of accuracy) and explainability, where highly accurate machine learning methods (such as deep-learning) are the least explainable, and the most explainable methods (such as decision trees) are the least accurate. In addition, explainability of artificial intelligence techniques in terms of transparency and post-hoc metrics are discussed. This report outlines the importance of data novelty and value of new information in evaluating both the explainability and trustworthiness. Novelty detection helps to establish consistency or inconsistency of the new data with respect to the training data. On the other hand, value of information could be a part of the user-centric visualization recommendation system that request additional information to be collected, thereby strengthening the machine learning outcomes. During this project, a copyrighted user-centric visualization that aligns with a human-in-the-loop approach was developed. The user-centric visualization presents different levels of information and can be tailored as per user credentials to gain user confidence. One of the salient features of the user-centric visualization is it presents machine learning methods with explainability metrics. A simplified version of the user-centric visualization was presented to 32 users with varying levels of machine learning expertise. Feedback was solicited to test the hypothesis that the app contained sufficient explainability and that the users would trust the algorithm. Overall, the app was positively received, and the hypothesis was supported. This report discusses the trust-but-verify framework – a potential approach to build user trust artificial intelligence. The framework discusses trust from the human level to artificial intelligence level. The fundamental premise of the trust but verify framework is derived from an observation of nuclear safety culture (i.e., nuclear power plant personnel do not rely on a singular source of data to make a decision). This also ties back to the user-centric visualization that presents different levels of information to achieve both explainability and trustworthiness of artificial intelligence. Even so, the adoption of artificial intelligence and machine learning in the nuclear industry faces additional barriers, namely regulatory and stakeholder readiness. To overcome these challenges, new solutions must gain regulatory approval and cater to stakeholder needs. The Nuclear Regulatory Committee has a 5-year strategic plan which prepares them for reviewing artificial intelligence technologies in licensee submissions. Early and frequent engagement with the regulator is encouraged. Additionally, artificial intelligence solutions should incorporate human-in-the-loop considerations and offer explainability. Stakeholders must prepare by hiring or training staff to adapt to advancing technology in everyday plant tasks.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

The Future of Robotics/Remote Handling Solutions for the Nuclear Industry - 17347

The nuclear industry is seeing a resurgence of development and utilization of advanced integrated systems; fused sensor arrays; autonomous, remote and hybrid operations; and augmented reality display and control system interfaces. A new age of 'robotics' and remote handling technologies. For many years the use of robotics in nuclear applications has been done sparingly due to technical limitations, perceived and real reliability issues, perceived and real risk concerns, and cost. Although robotics have been used successfully on a variety of projects globally, the underuse of this technology, particularly within the nuclear industry both public and private, is overwhelming. As the number of complex, high hazard projects in the nuclear industry increases, the use of robotics for maintenance, inspection, repair, decommissioning, and remediation activities will become increasingly vital in the future. The risks associated with direct human exposure are high and thus eliminating the risk, or managing it to acceptable levels by using robotics/remote handling as well as other technologies is critical to success as more extreme environment operations and tasks are undertaken. Recent and ongoing advances in technology now allows us to safely and effectively remove the direct / in-situ human element from these type of activities. All the while delivering the same or better outcomes when considering the alternatives when limited or no direct / in-situ human intervention would be allowed. The biggest obstacles the nuclear industry faces center on the cost, reliability, maintainability, effectiveness and appropriate uses of robotics/remote handling solutions in a nuclear (radiation and/or contamination) environment. In order to mitigate these concerns, a sound approach to the use of robotics needs to be addressed. This approach must include: 1. A better understanding of what technology is available and how it can be utilized across multiple sectors, e.g. Commercial Off-the-Shelf (COTS) technology modified / adapted for the nuclear sector; 2. How to implement technology in an integrated pragmatic approach that is cost effective by combining technological know-how / subject matter expertise (SME) with the equivalent subject matter expertise in regulatory, operational and maintenance requirements and limitations, such that a more complete set of functional and operational requirements can be jointly developed in the process. 3. The integration of requirements (project, site, regulatory, etc.), risk management (human factors) and technology, i.e. a marriage between political, programmatic, and technological factors. As the development of robotic/remote handling technology progresses and the successful implementation on projects is demonstrated, the better off the industry will be. This paper provides an overview of the fundamental issues facing the nuclear industry with regard to effective and successful technology selection, maturation, deployment, and integrated risk management, and offers suggestions on how we may overcome these persistent issues now and in the future. Robotic systems and solution experts that are focused on the integration of proven technologies - and that are an integrated part of the project team from its onset - are a necessity for successful planning during nuclear operations. The approach to nuclear operations should always be, 'what is the simplest, most effective way we can approach a particular project need?'. Understanding and implementing the right technologies is where the true value will be recognized. The integration of systems, which includes more than just robotics, is key when developing schemes for these types of complex operations. Often times during planning, the remote handling needs of the job are set aside to be evaluated at a later time due to risk concerns and lack of knowledge. The issue is, experts in the area of robotics or remote handling are rarely involved in the upfront planning. As a result, unnecessary constraints, risks, and obstacles are imposed on project teams to deliver. If done intelligently, this should be introduced at the forefront of planning. Using this type of sound approach leads to effective task organization for the end user. By developing an overall philosophy that leverages proven technologies with implementation expertise; i.e. solutions versus products, is a recipe for success. This is the area that is under-appreciated in the industry and one that needs to be recognized on a much broader scale. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Digital logic for soft devices

Although soft devices (grippers, actuators, and elementary robots) are rapidly becoming an integral part of the broad field of robotics, autonomy for completely soft devices has only begun to be developed. Adaptation of conventional systems of control to soft devices requires hard valves and electronic controls. This paper describes completely soft pneumatic digital logic gates having a physical scale appropriate for use with current (macroscopic) soft actuators. Each digital logic gate utilizes a single bistable valve—the pneumatic equivalent of a Schmitt trigger—which relies on the snap-through instability of a hemispherical membrane to kink internal tubes and operates with binary high/low input and output pressures. Soft, pneumatic NOT, AND, and OR digital logic gates—which generate known pneumatic outputs as a function of one, or multiple, pneumatic inputs—allow fabrication of digital logic circuits for a set–reset latch, two-bit shift register, leading-edge detector, digital-to-analog converter (DAC), and toggle switch. The DAC and toggle switch, in turn, can control and power a soft actuator (demonstrated using a pneu-net gripper). We note ehese macroscale soft digital logic gates are scalable to high volumes of airflow, do not consume power at steady state, and can be reconfigured to achieve multiple functionalities from a single design (including configurations that receive inputs from the environment and from human users). This work represents a step toward a strategy to develop autonomous control—one not involving an electronic interface or hard components—for soft devices.

36 MATERIALS SCIENCE↗

Conditional Karhunen–Loève regression model with Basis Adaptation for high-dimensional problems: Uncertainty quantification and inverse modeling

Here, we propose a methodology for improving the accuracy of surrogate models of the observable response of physical systems as a function of the systems’ spatially heterogeneous parameter fields, with applications to uncertainty quantification and parameter estimation in high-dimensional problems. Practitioners often formulate finite-dimensional representations of spatially heterogeneous parameter fields using truncated unconditional Karhunen–Loève expansions (KLEs) for a certain choice of unconditional covariance kernel and construct surrogate models of the observable response with respect to the KLE coefficients. When direct measurements of the parameter fields are available, we propose improving the accuracy of these surrogate models by representing the parameter fields via conditional Karhunen-Loève expansions (CKLEs). CKLEs are constructed by conditioning the covariance kernel of the unconditional expansion on the direct measurements of the parameter field via Gaussian process regression, and then truncating the corresponding KLE. We apply the proposed methodology to constructing surrogate models via the Basis Adaptation (BA) method of the stationary hydraulic head response, measured at spatially discrete observation locations, of a groundwater flow model of the Hanford Site, as a function of the 1000-dimensional representation of the model’s log-transmissivity field. We find that BA surrogate models of the hydraulic head based on CKLEs are more accurate than BA surrogate models based on unconditional expansions for forward uncertainty quantification tasks. Furthermore, we find that inverse estimates of the hydraulic transmissivity field computed using CKLE-based BA surrogate models are more accurate than those computed using unconditional BA surrogate models.

97 MATHEMATICS AND COMPUTING↗

PRIMED for the Future: Purposing Raw Intake for Machine Learning-Enabled Detection (Final Report)

The COVID-19 pandemic demonstrated how a novel, elusive, and diffuse biological threat can engender uncertainty and misinformation, and it underscored the need for flexible analytical modalities agnostic to the identity of biological material. Yet even before the pandemic, recognition of the limitations of the current, list-based approach, which focuses on known pathogens and biotoxins, and of the importance of agent-agnostic biodetection was growing within the biosecurity community. In a 2018 report on “Biodefense in the Age of Synthetic Biology,” for example, the National Academy of Sciences stated that “an overreliance on the Select Agent List is a systematic weakness affecting many aspects of the United States’ current biodefense mitigation capability”. More recently, a group of biodefense researchers proposed the identification and adoption of “bioagent-agnostic signatures (BASs)” as a way of detecting and characterizing not only existing agents but also novel ones, an approach they believe will “enable a more flexible and resilient biodefense posture”. Indeed, the future of biodetection requires us to begin developing novel analytics that can identify anomalies and/or characteristics that indicate a potential threat, whether known or unknown, without looking for a specific signature that has been identified previously. To assess potential threats more rapidly, it is critical to develop agnostic artificial intelligence (AI)/machine learning (ML) systems that can be employed for real-time assessment of the nature and source of a perturbation. Such systems should be multi scale and multi-dimensional, integrating sensor data from a range of biological, chemical, and physical application spaces. Emerging deep learning (DL) models demonstrate exceptional promise for identification of discriminatory features within multi-dimensional datasets. DL models have the capacity to recognize and encode highly complex patterns in a wide range of input data modalities, including images, text, and biological/chemical/physical spectra. As such, they can execute a wide range of assessments and determinations that have traditionally required a human operator. The promise of advances in DL is apparent in the realm of human health and medicine. DL models have been validated for evaluating a variety of clinical threats to human health in a range of contexts, including infection and cancer, and they demonstrated improved performance in predicting stroke relative to human neurologists in some categories of data. Continuously evolving advances in AI/ML are expected to support more efficient evaluation of raw sequence, spectroscopy, and spectrometry data. For instance, recent advances and deployment of large language models (LLM) such as Generative Pre training Transformer (GPT) and Bidirectional Encoder Representations from Transformers (BERT) have already motivated application of these models for biological function prediction. As frameworks such as LLMs become larger and more complex in their representations, their capacity to serve as pre-trained models that can be fine-tuned for biological/biodetection purposes will similarly be amplified. While existing and emerging AI/ML have found broad applicability and use cases in the clinical sciences, development for environmental evaluation and biodetection has been limited. Functionalizing such capabilities for this purpose requires an understanding of the existing technical landscape and how the respective tools and algorithms are currently being employed. This landscape awareness then allows an assessment of the current practical capabilities of existing models and the anticipated requirements and development efforts that will be needed to adapt available algorithms for biodetection applications relevant to DHS. Leveraging expertise in biodetection, ML, and operational biodetection, the effort described in this report is comprised of a systematic landscape assessment (Subtask 2.1), comparative evaluation (Subtask 2.2), and formulation of a value proposition (Subtask 2.3) for the prospect of ML-enabled, agnostic biodetection from raw, or minimally-processed, datasets.

59 BASIC BIOLOGICAL SCIENCES↗

Hybridizing Machine Learning and Physically-based Earth System Models to Improve Prediction of Multivariate Extreme Events (AI Exploration of Wildland Fire Prediction)

Focal Areas: This project responds to two focal areas identified in the DOE Call for AI4ESP White Papers: 1) Predictive modeling through the use of artificial intelligence (AI) techniques, and 2) insights gleaned from complex data using explainable AI and big data analytics. Science Challenge: Large wildland fires (hereafter wildfires) appearing as high-impact compound climate extreme events are closely related to hydroclimate and water cycle extremes that modulate surface fuel supply and combustibility. These compound events have multivariate climatic features (e.g., temperature, precipitation, relative humidity, wind, lightning) and societal drivers (e.g., forest management, land use change, human caused ignitions). Meanwhile, they induce strong feedbacks to the coupled atmosphere, biosphere, and hydrosphere by perturbing regional and global radiation budget as well as ecological, biogeochemical, and water cycles across multiple spatiotemporal scales. The nonlinear interactions between these natural and anthropogenic components of the Earth system are too complex to be completely and adequately represented in today’s Earth system models (ESMs). The inherent stochastic nature of fire activity at all scales further increases the difficulty of its prediction using ESMs that are usually developed from deterministic equations and parameterizations. Besides, concurrence of long-term (decadal to interdecadal) global climate change and fire regime shifts overlapping with short-term (intraseasonal to interannual) variations of regional fire weather and burning activity confound predictability of these compound extreme events. We propose to address the above scientific challenges by using machine learning (ML)-based data-driven modeling techniques to integrate observations and physically-based ESMs’ simulations in a computationally efficient hybrid prediction system. This prediction system is supposed to characterize the wildfire’s sensitivity to climate and exogenous drivers at high resolution (~ 0.25°) on subseasonal to seasonal (S2S) timescales providing improved predictability and explainability. We will use the system to help identify: (1) What are the computational elements of a hybrid system needed to predict compound climate extreme events such as global wildfires? (2) What are the key drivers (either natural or anthropogenic) that modulate short-term variations of multivariate fire weather and burning activity over different regions? How can one take advantage of those driver-response relationships to improve the predictability of large wildfires on S2S time scales? (3) What are the underlying physical mechanisms and sources of improved predictability? Which ML techniques are optimal in revealing and adapting these mechanisms?

54 ENVIRONMENTAL SCIENCES↗

Mondo: integrating disease terminology across communities

Precision medicine aims to enhance diagnosis, treatment, and prognosis by integrating multimodal data at the point of care. However, challenges arise due to the vast number of diseases, differing methods of classification, and conflicting terminological coding systems and practices used to represent molecular definitions of disease. This lack of interoperability artificially constrains the potential for diagnosis, clinical decision support, care outcome analysis, as well as data linkage across research domains to support the development or repurposing of therapeutics. There is a clear and pressing need for a unified system for managing disease entities⁠—including identifiers, synonyms, and definitions. To address these issues, we created the Mondo disease ontology—a community-driven, open-source, unified disease classification system that harmonizes diverse terminologies into a consistent, computable framework. Mondo integrates key medical and biomedical terminologies, including Online Mendelian Inheritance in Man (OMIM), Orphanet, Medical Subject Headings (MeSH), National Cancer Institute Thesaurus (NCIt), and more, to provide a comprehensive and accurate representation of disease concepts with fully provenanced and attributed links back to the sources. Mondo can be used as the handle for curation of gene–disease associations utilized in diagnostic applications, research applications such as computational phenotyping, and in clinical coding systems in clinical decision support by pointing the clinician to the numerous knowledge resources linked to the Mondo identifier. Mondo's community-centric approach, stewarded by the Monarch Initiative's expertise in ontologies, ensures that the ontology remains adaptable to the evolving needs of biomedical research and clinical communities, as well as the knowledge providers.

biomedical informatics↗

Towards philosophical reasoning with agentic LLMs: Socratic method for scientific assistance

As large language models (LLMs) become central tools in science, improving their reasoning capabilities is critical for meaningful and trustworthy applications. We introduce a Socratic agent for scientific reasoning, implemented through a structured system prompt that guides LLMs via classical principles of inquiry. Unlike typical prompt engineering or retrieval-based methods, our approach leverages definition, analogy, hypothesis elimination, and other Socratic techniques to generate more coherent, critical, and domain-aware responses. We evaluate the agent across diverse scientific domains and benchmark it on the abstraction and reasoning corpus challenge dataset, achieving 97.15% under a fixed prompting protocol and without fine-tuning or external tools. Expert evaluation shows improved reasoning depth, clarity, and adaptability over conventional LLM outputs, suggesting that structured prompting rooted in philosophical reasoning can improve the scientific utility of language models.

LLM reasoning↗

Nuclear Forensics: A Capability at Risk (Abbreviated Version)

Nuclear forensics is important to our national security. Actions, including provision of appropriate funding, are needed now to sustain and improve the nation's nuclear forensics capabilities. The Department of Homeland Security (DHS), working with cooperating agencies and national laboratories, should plan and implement a sustainable, effective nuclear forensics program. Nuclear forensics is the examination and evaluation of discovered or seized nuclear materials and devices or, in cases of nuclear explosions or radiological dispersals, of detonation signals and post-detonation debris. Nuclear forensic evidence helps law enforcement and intelligence agencies work toward preventing, mitigating, and attributing a nuclear or radiological incident. This report, requested by DHS, the National Nuclear Security Administration, and the Department of Defense, makes recommendations on how to sustain and improve U.S. nuclear forensics capabilities. The United States has developed a nuclear forensics capability that has been demonstrated in real-world incidents of interdicted materials and in exercises of actions required after a nuclear detonation. The committee, however, has concerns about the program and finds that without strong leadership, careful planning, and additional funds, these capabilities will decline. Major areas of concern include: Organization. The responsibility for nuclear forensics is shared by several agencies without central authority and with no consensus on strategic requirements to guide the program. This organizational complexity hampers the program and could prove to be a major hindrance operationally. Sustainability. The nation's current nuclear forensics capabilities are available primarily because the system of laboratories, equipment, and personnel upon which they depend was developed and funded by the nuclear weapons program. However, the weapons program's funds are declining. Workforce and Infrastructure. Personnel skilled in nuclear forensics are too few and are spread too thinly. Some key facilities are in need of replacement because they are old, outdated, and not built to modern environmental, health, and safety standards. Procedures and Tools. Most nuclear forensics techniques were developed to carry out Cold War missions and to satisfy a different, less restrictive set of environmental, health, and safety standards. Some of the equipment also does not reflect today's technical capabilities. The Executive Office of the President established the National Technical Nuclear Forensics Center under the direction of the Secretary of Homeland Security, to coordinate nuclear forensics in the United States. DHS's responsibility can only be carried out with the cooperation and support of the other agencies involved. The committee recommends that DHS and the other cooperating agencies should: 1. Streamline the organizational structure, aligning authority and responsibility; and develop and issue appropriate requirements documents. 2. Issue a coordinated and integrated implementation plan for fulfilling the requirements and sustaining and improving the program's capabilities. This plan would form the basis for the agencies' multi-year program budget requests. 3. Implement a plan to build and maintain an appropriately sized and composed nuclear forensics workforce, ensuring sufficient staffing at the national laboratories and support for university research, training programs, and collaborative relationships among the national laboratories and other organizations. 4. Adapt nuclear forensics to the challenges of real emergency situations, including, for example, conducting more realistic exercises that are unannounced and that challenge regulations and procedures followed in the normal work environment, and implementing lessons learned. The national laboratories should: 5. Optimize procedures and equipment through R&D to meet program requirements. Modeling and simulation should play an increased role in research, development, and planning. The nuclear forensics community should: 6. Develop standards and procedures for nuclear forensics that are rooted in the same underlying principles that have been recommended to guide modern forensic science. DHS and the other cooperating agencies should: 7. Devise and implement a plan that enables access to relevant information in databases including classified and proprietary databases for nuclear forensics missions. The Executive Office of the President and the Department of State, working with the community of nuclear forensics experts, should: 8. Determine the classes of data and methods that are to be shared internationally and explore mechanisms to accomplish that sharing.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

Large language models (LLMs) are increasingly adapted to downstream tasks via reinforcement learning (RL) methods like Group Relative Policy Optimization (GRPO), which often require thousands of rollouts to learn new tasks. We argue that the interpretable nature of language often provides a much richer learning medium for LLMs, compared to policy gradients derived from sparse, scalar rewards. To test this, we introduce GEPA (Genetic-Pareto), a prompt optimizer that thoroughly incorporates natural language reflection to learn high-level rules from trial and error. Given any AI system containing one or more LLM prompts, GEPA samples trajectories (e.g., reasoning, tool calls, and tool outputs) and reflects on them in natural language to diagnose problems, propose and test prompt updates, and combine complementary lessons from the Pareto frontier of its own attempts. As a result of GEPA's design, it can often turn even just a few rollouts into a large quality gain. Across six tasks, GEPA outperforms GRPO by 6% on average and by up to 20%, while using up to 35x fewer rollouts. GEPA also outperforms the leading prompt optimizer, MIPROv2, by over 10% (e.g., +12% accuracy on AIME-2025), and demonstrates promising results as an inference-time search strategy for code optimization. We release our code at https://github.com/gepa-ai/gepa.

97 MATHEMATICS AND COMPUTING↗