Search NASA⌕ Search

SEARCH · Search NASA

Results for “Association Learning”

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 667 records · Page 37
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

On Road vs. Off Road Low Load Cycle Comparison

Reducing criteria pollutants while reducing greenhouse gases is an active area of research for commercial on-road vehicles as well as for off-road machines. The heavy duty on-road sector has moved to reducing NOx by 82.5% compared to 2010 regulations while increasing the engine useful life from 435,000 to 650,000 miles by 2027 in the United States (US). An additional certification cycle, the Low Load Cycle (LLC), has been added focusing on part load operation having tight NOx emissions levels. In addition to NOx, the total CO2 emissions from the vehicle will also be reduced for various model years. The off-road market is following with a 90% NOx reduction target compared to Tier 4 Final for 130-560 kW engines along with greenhouse gas targets that are still being established. The off-road market will also need to certify with a Low Load Application Cycle (LLAC), a version of which was proposed for evaluation in 2021. Since the LLAC has not been finalized, this study is being conducted to compare and contrast the LLC for on-road with the LLAC for off-road as there might be some shared learnings. A US off-road production 2023 Fiat Powertrain 13L engine and aftertreatment system was chosen for this work. This engine is used in production for both off-road and on-road products, so it is a good choice for this study. The associated off-road aftertreatment system was aged for more relevant comparisons. The engine calibration was not altered for either of the low load cycles. This study shows that the cycles are quite different in nature as the market needs are different. The LLC includes a large fraction of operation at idle and lower speeds, representing products that use the engine primarily for motive power, where lower vehicle speed means a lower engine speed and load. The LLAC has more time and load spent at high speeds and slightly higher loads. The off-road products represented by this cycle often use the engine to drive auxiliary equipment which means higher parasitic loads and hand/fixed throttle. The comparison will include the use profiles, tailpipe NOx and greenhouse gas emissions (CO2, N2O).

McCarthy, James↗
DOE OSTI2026

Predictive models of the genetic bases underlying budding yeast fitness in multiple environments

Abstract The ability of organisms to adapt and survive depends on the effects of genes and the environment on fitness. However, the multigenic nature of fitness and genotype-by-environment interactions hinder our understanding of the genetic basis of fitness. Here, we established fitness prediction models for 35 environments using machine learning and existing fitness data and different genetic variant types for a Saccharomyces cerevisiae population. Models revealed that the predictive ability of genetic variants varied across environments, with copy number variants explaining the majority of fitness variation in most cases. Model interpretation showed that different variant types identified distinct gene sets associated with predictive variants. These gene sets were significantly enriched in experimentally validated genes affecting fitness in only a subset of environments, indicating that many genes influencing fitness remain unexplored. Notably, non-experimentally validated genes were more important than validated ones for fitness predictions. Gene contributions to predictions were both isolate- and environment-dependent, pointing to gene-by-gene and gene-by-environment interactions. Furthermore, models uncovered experimentally validated and novel candidate genetic interactions for a well-characterized stress, the fungicide benomyl. These findings highlight the feasibility of identifying the genetic basis of fitness by using different genetic variant types and offer novel targets for future functional analysis.

DNA copy number variations↗
DOE OSTI2025

Predicting Pulsed-Laser Deposition SrTiO 3 Homoepitaxy Growth Dynamics Using High-Speed Reflection High-Energy Electron Diffraction

Pulsed-laser deposition (PLD) is a powerful technique for growing complex oxides with controlled stoichiometry. To understand growth dynamics therein, it is common to leverage in situ spectroscopies, such as reflection high-energy electron diffraction (RHEED), to monitor surface crystallinity. Most commercial systems rely on video-rate cameras operating at 60-120 Hz that lack sufficient temporal resolution to capture growth dynamics at practical deposition frequencies. Here, a high-speed platform to record in situ dynamics via RHEED at >500 Hz is implemented. An open-source analysis package is designed to fit diffraction spots to 2D Gaussians, allowing single-pulse surface reconstruction kinetics extraction. Using homoepitaxially deposited (001)-oriented SrTiO 3 as a model system, we demonstrate how high-speed RHEED can provide real-time insight into growth processes obscured by slower acquisition systems. By fitting the single-pulse intensity to a set of exponential functions, we observe changes in the characteristic decay time and mechanism correlated to the substrate step width and surface termination. We observe distinct surface effects, with diffraction intensity decaying on lower-energy TiO 2 -terminated surfaces and stabilizing on SrO- or mixed-terminated surfaces. Similarly, using an exponential model, the extracted characteristic time of adatom deposition decreases with increased density of bonding sites associated with mixed termination and narrower step widths. Ultimately, this work shows how increasing RHEED temporal resolution can uncover new insights into growth processes, with practical implications for the design and control of PLD processes. This experimental platform provides new capabilities to enable data-driven machine learning analysis and autonomous control systems to enhance the complexity and fecundity of PLD.

(SrO)↗
DOE OSTI2024

Validation of new and existing methods for time-domain simulations of turbulence and loads

We seek to obtain a second-by-second match between the simulated and measured structural loads of a utility-scale wind turbine. To obtain the one-to-one load simulations, we start with the furthest upstream component of the modeling chain: the turbulent inflow. We consider new and existing methods to generate constrained-turbulence flow fields. The new method is based on large-eddy simulations (LES) and machine learning (ML). The existing methods include Kaimal-based TurbSim and the superstatistical wind field model. The inflow measurements used to constrain these simulations are obtained with a nacelle-mounted scanning lidar. We compare the flow fields for the different inflow simulation approaches and validate their associated load predictions against measurements collected in the Rotor Aero-dynamics, Aeroelastics, and Wake (RAAW) field campaign. We find that the rotor-position control developed for this study is key in enabling the time match between measurements and simulations. When this control approach is used, the load simulation performance tracks with the inflow simulation fidelity, with LES+ML yielding errors ≤ 4% for the damage-equivalent loads of flapwise bending moment, and tower fore-aft bending moments.

17 WIND ENERGY↗
NASA NTRS1983

Compiling the space shuttle wind tunnel data base: An exercise in technical and managerial innovators

Engineers evaluating Space Shuttle flight data and performance results are using a massive data base of wind tunnel test data. A wind tunnel test data base of the magnitude attained is a major accomplishment. The Apollo program spawned an automated wind tunnel data analysis system called SADSAC developed by the Chrysler Space Division. An improved version of this system renamed DATAMAN was used by Chrysler to document analyzed wind tunnel data and data bank the test data in standardized formats. These analysis documents, associated computer graphics and standard formatted data were disseminated nationwide to the Shuttle technical community. These outputs became the basis for substantiating and certifying the flight worthiness of the Space Shuttle and for improving future designs. As an aid to future programs this paper documents the lessons learned in compiling the massive wind tunnel test data base for developing the Space Shuttle. In particular, innovative managerial and technical concepts evolved in the course of conceiving and developing this successful DATAMAN system and the methods and organization for applying the system are presented.

Kemp, N. D.↗
NASA NTRS2016

NEEMO 20: Science Training, Operations, and Tool Development

The 20th mission of the National Aeronautics and Space Administration (NASA) Extreme Environment Mission Operations (NEEMO) was a highly integrated evaluation of operational protocols and tools designed to enable future exploration beyond low-Earth orbit. NEEMO 20 was conducted from the Aquarius habitat off the coast of Key Largo, FL in July 2015. The habitat and its surroundings provide a convincing analog for space exploration. A crew of six (comprised of astronauts, engineers, and habitat technicians) lived and worked in and around the unique underwater laboratory over a mission duration of 14-days. Incorporated into NEEMO 20 was a diverse Science Team (ST) comprised of geoscientists from the Astromaterials Research and Exploration Science (ARES/XI) Division from the Johnson Space Center (JSC), as well as marine scientists from the Department of Biological Sciences at Florida International University (FIU). This team trained the crew on the science to be conducted, defined sampling techniques and operational procedures, and planned and coordinated the science focused Extra Vehicular Activities (EVAs). The primary science objectives of NEEMO 20 was to study planetary sampling techniques and tools in partial gravity environments under realistic mission communication time delays and operational pressures. To facilitate these objectives two types of science sites were employed 1) geoscience sites with available rocks and regolith for testing sampling procedures and tools and, 2) marine science sites dedicated to specific research focused on assessing the photosynthetic capability of corals and their genetic connectivity between deep and shallow reefs. These marine sites and associated research objectives included deployment of handheld instrumentation, context descriptions, imaging, and sampling; thus acted as a suitable proxy for planetary surface exploration activities. This abstract briefly summarizes the scientific training, scientific operations, and tool development conducted during NEEMO 20 with an emphasis on the primary lessons learned.

Graff, T.↗
NASA NTRSDate not supplied

Recent Advances in Discrete Crack Modeling Applied to Laminated Composites with Emphasis on: Floating Node Method, VCCT and Cohesive Zone Modeling

The present talk will provide an overview of the work performed during the Advanced Composites Project (ACP) on the development, and verification and validation of the Floating Node Method (FNM) as well as the Virtual Crack Closure Technique (VCCT) and cohesive zone modeling (CZM). The FNM is a finite element-based technique to represent crack networks. The complex nature of matrix crack-delamination interactions observed in unidirectional (UD) tape laminates suggests that a methodology such as the FNM method may be required to be able to accurately simulate damage progression in these laminates. Simulating crack onset and growth within the context of the FNM relies on techniques such as VCCT and CZM. The talk is organized to provide, via select examples, an overview of the breadth of the Verification & Validation (V&V) exercises performed during the ACP, and how these challenged the state-of-the art and guided further developments in discrete crack modeling, while helping to establish confidence in the progress made and map the challenges ahead. The performance of the VCCT and CZM individually, and in combination with the FNM can be assessed through verification exercises. These exercises typically consist of a comparison of simulation results to known numerical or analytical solutions. Verification is key to identify implementation issues and limitations that, otherwise, may remain undetected and cloud any subsequent validation efforts. Indeed, a subset of these numerical exercises led to further developments of the VCCT and the FNM method as will be illustrated. Before embarking on the subsequent validation of the framework, it is critical to have adequate characterization data. However, the testing campaign conducted revealed material responses that challenged the state-of-the-art and required further developments. The developments in CZM technology associated with the modeling of the responses of hybrid interfaces (fabric/UD) will be given as an example. Finally, the talk will conclude with a summary of the validation exercises performed under quasi-static and fatigue loadings, highlighting some of the key achievements, outstanding challenges and lessons learned.

finite elements↗
DOE OSTI2024

W2VPCA: A Machine Learning Method for Measuring Attitudes With Natural Language

Company strategy influences many decisions in freight transportation. Behavioral models of company decision-making therefore could benefit from including strategy variables. However, strategy is difficult to observe and quantify. Attitudinal surveys of company executives can be used to collect measurements of latent strategy to use in quantitative models. However, surveys are costly and burdensome. Text mining methods to collect measurements overcome these issues somewhat, but typically require manual intervention and ignore the context of words, which can be problematic. This study introduces a new machine learning method to generate strategy measurement data from existing big text data. The new method, called W2VPCA, combines Natural Language Processing and Principal Components Analysis. W2VPCA produces measurement data that serve as quantitative indicators of latent strategy in behavioral models. W2VPCA is unsupervised, data-driven, and uses information on word context. We apply W2VPCA to generate measurements of latent strategies using readily available, large-scale text data: annual company reports. The empirical measurements are used successfully to associate two latent strategies, one focusing on distribution and the other on products, with truck fleet and distribution center outsourcing decisions. The main empirical outcome is that the W2VPCA measurements outperform Bag-of-Words measurements in a psychometric analysis of latent firm strategies. While this study focuses on freight behavioral models, W2VPCA may also have applications in behavioral modeling in other domains.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Coarse-graining Hamiltonian systems using WSINDy

Abstract Weak form equation learning and surrogate modeling has proven to be computationally efficient and robust to measurement noise in a wide range of applications including ODE, PDE, and SDE discovery, as well as in coarse-graining applications, such as homogenization and mean-field descriptions of interacting particle systems. In this work we extend this coarse-graining capability to the setting of Hamiltonian dynamics which possess approximate symmetries associated with timescale separation. A smooth $$\varepsilon$$ ε -dependent Hamiltonian vector field $$X_\varepsilon$$ X ε possesses an approximate symmetry if the limiting vector field $$X_0=\lim _{\varepsilon \rightarrow 0}X_\varepsilon$$ X 0 = lim ε → 0 X ε possesses an exact symmetry. Such approximate symmetries often lead to the existence of a Hamiltonian system of reduced dimension that may be used to efficiently capture the dynamics of the symmetry-invariant dependent variables. Deriving such reduced systems, or approximating them numerically, is an ongoing challenge. We demonstrate that WSINDy can successfully identify this reduced Hamiltonian system in the presence of large perturbations imparted in the $$\varepsilon >0$$ ε > 0 regime, while remaining robust to extrinsic noise. This is significant in part due to the nontrivial means by which such systems are derived analytically. WSINDy naturally preserves the Hamiltonian structure by restricting to a trial basis of Hamiltonian vector fields. The methodology is computationally efficient, often requiring only a single trajectory to learn the global reduced Hamiltonian, and avoiding forward solves in the learning process. In this way, we argue that weak-form equation learning is particularly well-suited for Hamiltonian coarse-graining. Using nearly-periodic Hamiltonian systems as a prototypical class of systems with approximate symmetries, we show that WSINDy robustly identifies the correct leading-order system, with dimension reduced by at least two, upon observation of the relevant degrees of freedom. While our main contribution is computational, we also provide a contribution to the literature on averaging theory by proving that first-order averaging at the level of vector fields preserves Hamiltonian structure in nearly-periodic Hamiltonian systems. This provides theoretical justification for our approach as WSINDy’s computations occur at the level of Hamiltonian vector fields. We illustrate the efficacy of our proposed method using physically relevant examples, including coupled oscillator dynamics, the Hénon–Heiles system for stellar motion within a galaxy, and the dynamics of charged particles.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Machine Learning for Joint Quality Control

The use of lightweight material combinations has been highly demanded in manufacturing automotive structures. However, making robust dissimilar material joints of such lightweight materials is still challenging. A significant barrier to achieving high-quality and repeatable joint performance is a deficient understanding of the relationship between the welding process, joint attributes, and joint performance. In this context, welding factors refer to material, equipment, environment, and process parameters, while joint features comprise specific microstructural attributes of the weld such as nugget size, heat affected zone (HAZ) topology, intermetallic layer thickness, and sheet thickness reduction. Joint performance is quantified in terms of strength (e.g., tensile shear, coach peel, cross-tension), weld size, and hardness, among other factors. While there have been many attempts to establish this process-structure-property relationship by developing a model derived from the associated physics and first principles, the complexity of the joining processes compounded by the complex interactions with different materials in an automotive assembly line environment, has hindered the usefulness of such attempts. The complexity is further exacerbated using different stacking materials, especially comprising dissimilar material combinations. In practice, the common approach has been the laborious process of creating welds, characterizing them, and then physically testing them through experimentation. With the emergence of artificial intelligence (AI) methods, an alternative pathway to eliciting the desired process-structure-property relationship at an accelerated pace is to use a data-driven approach by employing machine-learning (ML) techniques. This approach is benefitted by the availability of large streams of data, generated through years of research and testing by original equipment manufacturers, in the form of material, process, environmental, equipment, microstructural, and bulk-scale performance information from multimodal, multiscale sensors making measurements from laboratory-scale to production-scale processes. During Phase I efforts, which ended in fiscal year (FY) 2021, the Oak Ridge National Laboratory and Pacific Northwest National Laboratory (ORNL/PNNL) team demonstrated the effectiveness of different ML/AI frameworks in modeling complex relationships between resistance spot welding (RSW) process parameters, weld attributes, and joint properties using a subset of data from General Motors (GM). In FY 2022, the project team further refined and expanded their respective ML models to analyze additional welds with new weld stack-ups and materials to enhance the ML model predictive capability. ORNL extended its unified deep neural networks (DNN) ML training and prediction framework with new data streams of process parameters, and PNNL extended its model describing RSW process parameters’ associations with weld attributes. In FY 2023, the project team completed the development of the AI/ML architecture for analyzing aluminum/steel joints manufactured by GM via RSW and transitioned into the inline welding quality monitoring task for steel/steel RSW joints provided by GM.

36 MATERIALS SCIENCE↗
DOE OSTI2024

KRT18 as a Novel Biomarker of Urothelial Papilloma while Evaluating Low-Grade Papillary Urothelial Neoplasms: Bi-Center Analysis

Introduction: Although urothelial papilloma (UP) is an indolent papillary neoplasm that can mimic the morphology of low-grade papillary urothelial carcinoma (PUC), there is no immunomarker to differentiate reliably these two entities. In addition, the molecular characteristics of UP are not fully understood. Methods: We conducted an in-depth proteomic analysis of papillary urothelial lesions (n = 31), including UP and PUC along with normal urothelium. Protein markers distinguishing UP and PUC were selected with machine learning analysis, followed by internal and external validation using immunohistochemistry. Results: In the proteomic analysis, UP and PUC showed overlapping proteomic profiles. Here, we identified EHD4 and KRT18 as candidate diagnostic biomarkers of UP. Through immunohistochemical validation in two independent cohorts (n = 120), KRT18 was suggested as a novel UP diagnostic marker, able to differentiate UP from low-grade PUC. We also found that 3.5% of patients with UP developed urothelial carcinoma in subsequent resections, supporting the malignant potential of UP. KRT18 downregulation was significantly associated with UPs subsequently progressing to urothelial carcinoma, following their initial diagnosis. Conclusion: This is the first study that successfully revealed UPs comprehensive proteomic landscape, while it also identified KRT18 as a potential diagnostic biomarker of UP.

Biomarkers↗
NASA NTRS1994

Empirical modeling for intelligent, real-time manufacture control

Artificial neural systems (ANS), also known as neural networks, are an attempt to develop computer systems that emulate the neural reasoning behavior of biological neural systems (e.g. the human brain). As such, they are loosely based on biological neural networks. The ANS consists of a series of nodes (neurons) and weighted connections (axons) that, when presented with a specific input pattern, can associate specific output patterns. It is essentially a highly complex, nonlinear, mathematical relationship or transform. These constructs have two significant properties that have proven useful to the authors in signal processing and process modeling: noise tolerance and complex pattern recognition. Specifically, the authors have developed a new network learning algorithm that has resulted in the successful application of ANS's to high speed signal processing and to developing models of highly complex processes. Two of the applications, the Weld Bead Geometry Control System and the Welding Penetration Monitoring System, are discussed in the body of this paper.

Xu, Xiaoshu↗
NASA NTRS2008

Architecture for Multiple Interacting Robot Intelligences

An architecture for robot intelligence enables a robot to learn new behaviors and create new behavior sequences autonomously and interact with a dynamically changing environment. Sensory information is mapped onto a Sensory Ego-Sphere (SES) that rapidly identifies important changes in the environment and functions much like short term memory. Behaviors are stored in a database associative memory (DBAM) that creates an active map from the robot's current state to a goal state and functions much like long term memory. A dream state converts recent activities stored in the SES and creates or modifies behaviors in the DBAM.

Peters, Richard Alan, II↗
NASA NTRS2010

Definition and Demonstration of a Methodology for Validating Aircraft Trajectory Predictors

This paper presents a new methodology for validating an aircraft trajectory predictor, inspired by the lessons learned from a number of field trials, flight tests and simulation experiments for the development of trajectory-predictor-based automation. The methodology introduces new techniques and a new multi-staged approach to reduce the effort in identifying and resolving validation failures, avoiding the potentially large costs associated with failures during a single-stage, pass/fail approach. As a case study, the validation effort performed by the Federal Aviation Administration for its En Route Automation Modernization (ERAM) system is analyzed to illustrate the real-world applicability of this methodology. During this validation effort, ERAM initially failed to achieve six of its eight requirements associated with trajectory prediction and conflict probe. The ERAM validation issues have since been addressed, but to illustrate how the methodology could have benefited the FAA effort, additional techniques are presented that could have been used to resolve some of these issues. Using data from the ERAM validation effort, it is demonstrated that these new techniques could have identified trajectory prediction error sources that contributed to several of the unmet ERAM requirements.

Vivona, Robert A.↗
NASA NTRSDate not supplied

Adaptative Site Management for a 115 Acre Chlorinated Solvent Plume with Two Separate Source Areas at Kennedy Space Center, Florida

Background/Objectives. During Resource Conservation and Recovery Act (RCRA) Facility Investigation (RFI) activities, Geosyntec delineated a chlorinated volatile organic compound (CVOC) plume at the National Aeronautics and Space Administration’s (NASA’s) Vehicle Assembly Building (VAB) area located at KSC, Florida. The RFI activities identified an approximate 115-acre dissolved plume (primarily vinyl chloride) and a trichloroethene (TCE) source area in an active aerospace complex that is surrounded by sensitive wetland/waterbodies. Due to the size of the impacted area, the Corrective Measure Design included a multi-component strategy: (i) address the source area via bioremediation; (ii) protect sensitive wetlands from impacted groundwater discharge via biosparging; and (iii) Long Term Monitoring (LTM) of the remaining dissolved plume. After the Corrective Measures implementation (CMI), NASA and Geosyntec worked with Florida Department of Environmental Protection (FDEP) to implement an adaptive site management for the complex, 115-acre site outside of the traditional RCRA process. The adaptive site management approach relied on performing supplemental assessments and implementing Interim Measures (IMs) to further assess and implement remedies over time while working within site and budget constraints, with an overall goal of achieving enough mass reduction to transition the entire site to LTM and eventually achieve site closure. Approach/Activities. After the biosparge barrier was operational and bioremediation within the source area (referred to as Hot Spot 1) achieved the Corrective Action Objective (CAO), supplemental assessment of the area between Hot Spot 1 and the biosparge barrier was performed. The conceptual site model was updated using the supplemental assessment results and an air sparge system IM was designed to treat an approximate 1.2 acre area (referred to as Hot Spot 2). After installation of the air sparge system, supplemental assessment within the remainder of the 115-acre dissolved plume was performed and a second TCE source area was identified. The TCE source area and associated areas with elevated CVOC concentrations (referred to as Hot Spot 3) were delineated and a bioremediation IM was implemented. Also, the downgradient impacts from Hot Spot 3 were adjacent to a sensitive waterbody, and negotiations with the FDEP allowed the area to be monitored using LTM. Results/Lessons Learned. The performance of supplemental assessment activities and implementation of remedial alternatives as IMs allowed NASA to successfully address groundwater impacts over time, while working within the FDEP regulatory framework. The implementation of the CMI and multiple IMs has achieved the following goals: (i) the biosparge barrier has mitigated the potential discharge of impacted groundwater to an adjacent wetland; (ii) enhanced bioremediation within Hot Spot 1 achieved the CAO within 2 years and transitioned the area into LTM; (iii) operation of an air sparge system within Hot Spot 2 removed TCE as a constituent of concern and contributed to a reduction (approximately 43%) in the impacted groundwater area outside the air sparge treatment area (plume collapse); and (iv) bioremediation within Hot Spot 3 removed approximately 80% of the CVOC mass and contributed to a reduction (approximately 47%) in the impacted groundwater area outside the bioremediation IM treatment area. Overall, the adaptive approach is protecting the sensitive water bodies surrounding the complex site and reducing the area of impacted groundwater, which is moving the entire site towards LTM.

Rebecca C Daprato↗
DOE OSTI2025

Systems and methods for binary code analysis

Human-readable (HR) code may be derived from a binary. The HR code may be configured to have statistical properties suitable for machine-learned (ML) translation. The HR code may comprise source code, intermediate code, assembly code, or the like. A machine-learned translator may be configured to translate the HR code into labels comprising semantic information pertaining to respective functions of the binary, such as a function name, role, or the like. Execution of the binary may be blocked in response to translating the HR code to a label associated with malware, such as cryptocurrency mining malware or the like. Conversely, the binary may be permitted to proceed to execution in response to determining that the translation is free from labels indicative of malware.

Anderson, Matthew W.↗
DOE OSTI2024

Toward Accelerating Discovery via Physics-Driven and Interactive Multifidelity Bayesian Optimization

Both computational and experimental material discovery bring forth the challenge of exploring multidimensional and often nondifferentiable parameter spaces, such as phase diagrams of Hamiltonians with multiple interactions, composition spaces of combinatorial libraries, processing spaces, and molecular embedding spaces. Often these systems are expensive or time consuming to evaluate a single instance, and hence classical approaches based on exhaustive grid or random search are too data intensive. This resulted in strong interest toward active learning methods such as Bayesian optimization (BO) where the adaptive exploration occurs based on human learning (discovery) objective. However, classical BO is based on a predefined optimization target, and policies balancing exploration and exploitation are purely data driven. In practical settings, the domain expert can pose prior knowledge of the system in the form of partially known physics laws and exploration policies often vary during the experiment. Here, we propose an interactive workflow building on multifidelity BO (MFBO), starting with classical (data-driven) MFBO, then expand to a proposed structured (physics-driven) structured MFBO (sMFBO), and finally extend it to allow human-in-the-loop interactive interactive MFBO (iMFBO) workflows for adaptive and domain expert aligned exploration. These approaches are demonstrated over highly nonsmooth multifidelity simulation data generated from an Ising model, considering spin–spin interaction as parameter space, lattice sizes as fidelity spaces, and the objective as maximizing heat capacity. Detailed analysis and comparison show the impact of physics knowledge injection and real-time human decisions for improved exploration with increased alignment to ground truth. Here, the associated notebooks allow to reproduce the reported analyses and apply them to other systems.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2011

The Story Behind the Numbers: Lessons Learned from the Integration of Monitoring Resources in Addressing an ISS Water Quality Anomaly

Beginning in June of 2010 an environmental mystery was unfolding on the International Space Station (ISS). The U.S. Water Processor Assembly (WPA) began to produce water with increasing levels of total organic carbon (TOC). A surprisingly consistent upward TOC trend was observed through weekly in-flight total organic carbon analyzer (TOCA) monitoring. As TOC is a general organics indicator, return of water archive samples was needed to make better-informed crew health decisions and to aid in WPA troubleshooting. TOCA-measured TOC was more than halfway to its health-based screening limit before archive samples could be returned on Soyuz 22 and analyzed. Although TOC was confirmed to be elevated, somewhat surprisingly, none of the typical target compounds were the source. After some solid detective work, it was confirmed that the TOC was associated with a compound known as dimethylsilanediol (DMSD). DMSD is believed to be a breakdown product of silicon-containing compounds present on ISS. A toxicological limit was set for DMSD and a forward plan developed for operations given this new understanding of the source of the TOC. This required extensive coordination with ISS stakeholders and innovative use of available in-flight and archive monitoring resources. Behind the numbers and scientific detail surrounding this anomaly, there exists a compelling story of multi-disciplinary awareness, teamwork, and important environmental lessons learned.

McCoy, Torin↗