Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Continual 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 1,369 records · Page 76

On-demand Command and Control of ASTERIA with Cloud-based Ground Station Services

ASTERIA (Arcsecond Space Telescope Enabling Research in Astrophysics) was a 6-unit CubeSat technology demonstration mission that deployed from the International Space Station on November 20th, 2017. After successfully completing its 90-day primary mission that demonstrated arcsecond-level line-of-sight pointing and focal plane thermal stability for exoplanet detection, it entered an extended mission performing onboard software demonstrations to mature technology both in space and on the ground. One of the technologies was a completely cloud-based ground system leveraging Amazon Web Services (AWS) Ground Station service. Announced in December 2018 and launched in May 2019, AWS Ground Station is a fully managed ground station service that aims to reduce the overhead associated with developing and maintaining ground system infrastructure throughout the mission lifecycle. AWS Ground Station makes available the suite of features required for any ground system in support of low-Earth orbit (LEO) and medium-Earth Orbit (MEO) satellite operations on-demand and without setting up or maintaining long-term contracts. Charges are incurred on a per-minute basis for antenna usage during scheduled tracks. Support is available for S-band uplink and downlink, along with X-band narrowband and wideband downlink. Missions that use the service may reserve tracks with any licensed AWS Ground Station antennas located across each service region and have direct access to any AWS services in support of mission operations. The cloud-based architecture built around the AWS Ground Station service greatly enhanced ASTERIA mission operations by enabling end-to-end pass automation, on-demand contact scheduling and contingency planning, along with more efficient data downlink through station availability and station-tostation handovers. It incorporated open-source software, particularly NASA's AMMOS Instrument Toolkit (AIT) and Open Mission Control Technologies (OpenMCT), along with the AWS application programming interfaces (API) to the Ground Station, Elastic Compute Cloud (EC2) and Simple Storage Service (S3) services. After showcasing operability in August 2019, the team continued using and improving this novel ground system architecture until the end of mission in December 2019. This paper describes the cloud-based ground system, how it was designed, tested, and evaluated with an inorbit spacecraft, the operational capabilities that it enabled, along with lessons learned and recommendations for future missions.

Fesq, Lorraine↗

Learning to Embrace Nuclear Physics through Education

Due to its achievements, nuclear physics is more and more present in life of every member of the society. Its applications in the medical field and in nuclear energy, as well as the advanced research, always pushing the limits of science towards micro cosmos and macro cosmos, are subjects frequently presented in the media. In addition to their invaluable benefits, these achievements involve also particular rules to prevent potential risks. These risks are also underlined by the media, often being presented in an unfriendly manner. Specialists in nuclear physics are familiar with these problems complying with the specific rules in order to reduce risks at insignificant levels. The development of a specific field ('Radiation protection') defining norms and requirements for 'assuring the radiological safety of the workers, population and environment', and its dynamics represent a proof of a responsible attitude regarding nuclear safety. Dedicated international bodies and experts analyze and rigorously evaluate risks in order to draw the right ways of managing activity in the field. The improvement of the formal and informal education of public regarding the real risks of nuclear applications is very important in order to understand and better assimilate some general rules concerning the use of these techniques, as well as for their correct perception, leading to an increase of interest towards nuclear physics. This educational update can be started even from elementary school and continued in each stage of formal education in adapted forms. The task of informing general public is to be carried out mainly by specialists who, unlike 30-40 years ago, can rely on a much more efficient generation of communications' mean. Taking into account the lack of interest for nuclear, an attractive way of presenting the achievements and future possibilities of nuclear physics would contribute to youth orientation towards specific universities in order to become next generation of specialists in the field. Facing new challenges, society becomes aware of the fact that education represents the real solution to escalade them. Nuclear physics plays an important role in ensuring energetic resources for the near future and in reducing greenhouse effects. On the other hand, especially nuclear physics will permit to solve the enigma of universe birth. As in any other field, development involves continuous education and knowledge upgrading for all categories carrying out nuclear activities. For radiation protection workers and specialists, periodically refreshment courses are mandatory, in compliance with the national and international specific requirements.

99 GENERAL AND MISCELLANEOUS↗

Machine learning changes the rules for flux limiters

Learning to integrate non-linear equations from highly resolved direct numerical simulations has seen recent interest for reducing the computational load for fluid simulations. Here, we focus on determining a flux-limiter for shock capturing methods. Focusing on flux limiters provides a specific plug-and-play component for existing numerical methods. Since their introduction, an array of flux limiters has been designed. Using the coarse-grained Burgers' equation, we show that flux-limiters may be rank-ordered in terms of their log-error relative to high-resolution data. We then develop a theory to find an optimal flux-limiter and present flux-limiters that outperform others tested for integrating Burgers' equation on lattices with [Formula: see text], and 2x, 3x, 4x, and 8x coarse-grainings. We train a continuous piecewise linear limiter by minimizing the mean-squared misfit to six-grid point segments of high-resolution data, averaged over all segments. While flux limiters are generally designed to have an output of φ(r) = 1 at a flux ratio of r = 1, our limiters are not bound by this rule and yet produce a smaller error than standard limiters. Here we find that our machine learned limiters have distinctive features that may provide new rules-of-thumb for the development of improved limiters. Additionally, we use our theory to learn flux-limiters that outperform standard limiters across a range of values (as opposed to at a specific fixed value) of coarse-graining, number of discretized bins, and diffusion parameter. This demonstrates the ability to produce flux limiters that should be more broadly useful than standard limiters for general applications.

97 MATHEMATICS AND COMPUTING↗

Side chain engineering in indacenodithiophene- co -benzothiadiazole and its impact on mixed ionic–electronic transport properties

Organic semiconductors are increasingly being decorated with hydrophilic solubilising chains to create materials that can function as mixed ionic–electronic conductors, which are promising candidates for interfacing biological systems with organic electronics. While numerous organic semiconductors, including p- and n-type materials, small molecules and polymers, have been successfully tailored to encompass mixed conduction properties, common to all these systems is that they have been semicrystalline materials. Here, we explore how side chain engineering in the nano-crystalline indacenodithiophene-co-benzothiadiazole (IDTBT) polymer can be used to instil ionic transport properties and how this in turn influences the electronic transport properties. This allows us to ultimately assess the mixed ionic–electronic transport properties of these new IDTBT polymers using the organic electrochemical transistor as the testing platform. Using a complementary experimental and computational approach, we find that polar IDTBT derivatives can be infiltrated by water and solvated ions, they can be electrochemically doped efficiently in aqueous electrolyte with fast doping kinetics, and upon aqueous swelling there is no deterioration of the close interchain contacts that are vital for efficient charge transport in the IDTBT system. Despite these promising attributes, mixed ionic–electronic charge transport properties are surprisingly poor in all the polar IDTBT derivatives. Albeit a “negative” result, this finding clearly contradicts established side chain engineering rules for mixed ionic–electronic conductors, which motivated our continued investigation of this system. We eventually find this anomalous behaviour to be caused by increasing energetic disorder in the polymers with increasing polar side chain content. We have investigated computationally how the polar side chain motifs contribute to this detrimental energetic inhomogeneity and ultimately use the learnings to propose new molecular design criteria for side chains that can facilitate ion transport without impeding electronic transport.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Spatiotemporal forecasting of the edge localized modes in tokamak plasmas using neural networks

Artificial intelligence techniques have been increasingly adopted by the plasma and fusion science to address problems like plasma reconstruction, surrogate modeling, and tokamak/stellarator optimization. A key focus in sustained fusion research is the prediction and mitigation of edge-localized-modes (ELMs), instabilities that occur in short, periodic bursts and can cause erosion to the tokamak vessel wall. Recent research has demonstrated the power of neural networks in approximating continuous functions. In this work, we build spatiotemporal forecasting models that can predict the onset of ELMs and their evolution at early stages. We leverage recent advances in generative modeling, sequence-to-sequence modeling, and Fourier neural operators to propose architectures and training strategies that can learn to forecast short to long term dynamics of the noisy signals due to ELMs. We benchmark the developed model against a state-of-the-art foundation model using the beam emission spectroscopy (BES) data that captures the plasma fluctuations due to ELMs over a 8 x 8 spatial grid. Our models demonstrate high accuracy, outperforming the baselines, in predicting the evolution of BES signals during ELM events. Furthermore, the developed models exhibit high accuracy in predicting the rapid rise and relaxation of the signals due to ELMs within 30–80 µs.

edge localized modes↗

Learning-Based Inversion-Free Model-Data Integration to Advance Ecosystem Model Prediction

Ecosystem model prediction is important for understanding ecosystem responses to climate change and for management support. Model prediction and quantification of predictive uncertainty of ecosystems have long been investigated. The traditional workflow, which calibrates models to match observations and then uses the calibrated models for predictions, relies heavily on inverse modeling to constrain uncertain parameters in complex forward models. This inversion-based prediction approach is infeasible for complex models with heterogeneous parameter uncertainties and incapable of rapid integration of streaming and multiple sources of data because of the difficulty and computational cost in the model inversion, which is typically ill-posed and can require hundreds of thousands of expensive forward simulations to be performed iteratively. We propose to circumvent inverse modeling by precomputing an ensemble of unconstrained forward simulations and then using machine learning (ML) methods to learn the statistical relationship between simulated observation and prediction quantities. Once the ML model has learned the relationship, it can be used to make predictions of future system behavior with uncertainty quantification based on observations. The proposed learningbased inversion-free model prediction (LIMP) framework is computationally efficient which only requires a few thousands of fully parallelizable forward simulations. Additionally, LIMP can continually update predictions based on streaming observations from multiple locations and sources without necessarily requiring extra model simulations. In this study, we apply LIMP to a regional terrestrial ecosystem model with 47 parameters for testing, refining, and evaluating the approach. We demonstrate that LIMP can be used for efficient model prediction, rapid data assimilation, and cost-effective experimental design for improving robust predictive understanding of ecosystems.

Lu, Dan↗

Developing Expert Teams with a Strong Safety Culture

Would you like to lead a world renowned team that draws out all the talents and expertise of its members and consistently out performs all others in the industry? Ever wonder why so many organizations fail to truly learn from past mistakes only to repeat the same ones at a later date? Are you a program/project manager or team member in a high-risk organization where the decisions made often carry the highest of consequences? Leadership, communication, team building, critical decision-making and continuous team improvement skills and behaviors are mere talking points without the attitudes, commitment and strategies necessary to make them the very fabric of a team. Developing Expert Teams with a Strong Safety Culture, will provide you with proven knowledge and strategies to take your team soaring to heights you may have not thought possible. A myriad of teams have applied these strategies and techniques within their organization team environments: military and commercial aviation, astronaut flight crews, Shuttle flight controllers, members of the Space Shuttle Program Mission Management Team, air traffic controllers, nuclear power control teams, surgical teams, and the fire service report having spectacular success. Many industry leaders are beginning to realize that although the circumstances and environments of these teams may differ greatly to their own, the core elements, governing principles and dynamics involved in managing and building a stellar safety conscious team remain identical.

Rogers, David G.↗

Fostering Better Collaboration in Software Development Cycles Between Scientists and Programmers to Ensure the Integrity of and Promote the Development of New Scientific Data Products.

Misaligned incentives lead to reduced interaction between scientists and programmers on modern NASA science data-product development teams. Typically, situations arise where the scientist is not incentivized to learn modern coding practices and the programmer does not understand the science algorithms in the code. A programmer is responsible for the deliverable code thus setting a tradeoff between the desire for code improvement versus fear of compromising the integrity of data-product while the scientist continues to rely on their legacy codebases owing to the complexity of using the delivered code outside the processing environment and lack of validation modules. The NASA/CERES-TISA project has adopted a collaborative approach, with scientists and programmers both utilizing the same software repository with multiple branches, some optimized for delivery to a processing datacenter and others for scientific product development and validation. A team of scientists and programmers jointly review any new science code updates for integration into the codebase and strive to improve practices through promoting algorithm understanding, better institutional knowledge exchange and documentation, modularization, and developing data processing flow-dictated validation and debugging methods. This leads to a reduction in the personnel single point failures and reduced development time for creation of new science data-products.

CERES↗

Machine learning-based ethylene concentration estimation, real-time optimization and feedback control of an experimental electrochemical reactor

With the increase in electricity supply from clean energy sources, electrochemical reduction of carbon dioxide (CO 2 ) has received increasing attention as an alternative source of carbon-based fuels. As CO 2 reduction is becoming a stronger alternative for the clean production of chemicals, the need to model, optimize and control the electrochemical reduction of the CO 2 process becomes inevitable. However, on one hand, a first-principles model to represent the electrochemical CO 2 reduction has not been fully developed yet because of the complexity of its reaction mechanism, which makes it challenging to define a precise state-space model for the control system. On the other hand, the unavailability of efficient concentration measurement sensors continues to challenge our ability to develop feedback control systems. Gas chromatography (GC) is the most common equipment for monitoring the gas product composition, but it requires a period of time to analyze the sample, which means that GC can provide only delayed measurements during the operation. Moreover, the electrochemical CO 2 reduction process is catalyzed by a fast-deactivating copper catalyst and undergoes a selectivity shift from the product-of-interest at the later stages of experiments, which can pose a challenge for conventional control methods. To this end, machine learning (ML) techniques provide a potential approach to overcome those difficulties due to their demonstrated ability to capture the dynamic behavior of a chemical process from data. Motivated by the above considerations, we propose a machine learning-based modeling methodology that integrates support vector regression and first-principles modeling to capture the dynamic behavior of an experimental electrochemical reactor; this model, together with limited gas chromatography measurements, is employed to predict the evolution of gas-phase ethylene concentration. The model prediction is directly used in a proportional-integral (PI) controller that manipulates the applied potential to regulate the gas-phase ethylene concentration at energy-optimal set-point values computed by a real-time process optimizer (RTO). Specifically, the RTO calculates the operation set-point by solving an optimization problem to maximize the economic benefit of the reactor. Finally, suitable compensation methods are introduced to further account for the experimental uncertainties and handle catalyst deactivation. The proposed modeling, optimization, and control approaches are the first demonstration of active control for a CO 2 electrolyzer and contribute to the automation and scale-up efforts for electrified manufacturing of fuels and chemicals starting from CO 2 .

42 ENGINEERING↗

Machine Learning Applications to Metal-Silicate Equilibria and their Insights into Core Formation

An extensive number of studies have experimentally investigated how elements distribute between metal and silicate phases, to better constrain core-mantle chemical equilibrium. Here, we present a new database compiling all (to our knowledge) experimental data on liquid metal-silicate partitioning from 118 peer-reviewed publications. We applied various machine learning techniques to gain further insights into these partitioning equilibria and their dependencies. We performed a network analysis to investigate the relationship between experiments and partition coefficients, which enables visualizing gaps in the experimental dataset and biases related to varying experimental conditions and analytical setup. In addition, semi-empirical thermodynamic models are commonly used to extrapolate these chemical reactions to the wide range of pressure, temperature and compositional conditions of planetary differentiation. These models are based on linear regressions that assume continuous relationship between partition coefficients and experimental variables. Here, we considered random forest regressions, which are algorithms based on ensembles of decision trees and does not consider continuous effects of each variable. The application of this regression significantly improves the prediction of metal-silicate partitioning for several elements including Ni, Si and Cr. We will show how this new approach improves our understanding of elemental exchange between metal and silicate and their implications for the Earth’s core formation.

siderophile element↗

Smart Methane Emission Detection System Development (Final Report)

Working with the Department of Energy’s National Energy Technology Laboratory, Southwest Research Institute® (SwRI®) developed a system to identify methane leaks reliably, accurately, and autonomously at critical midstream sections of the natural gas distribution network in real- time for the purpose of mitigating methane emissions using Optical Gas Imaging (OGI) cameras. SwRI’s Smart Leak Detection – Methane (SLED/M) adds a high degree of automation to the process of methane leak detection to minimize sources of human error, minimize response time to a leak event, and maximize midstream visibility. Furthermore, SwRI has been working towards integrating Quantitative OGI (QOGI) capabilities into this existing technology. By leveraging Deep Learning, SwRI now has the capability to estimate fugitive emission leak rates quickly and reliably, which allows operators to detect emissions, quantify leak rate, prioritize repairs, and validate the repairs in a single instrument. The next generation QOGI technology leverages the same cameras used in Leak Detection and Repair (LDAR) programs, with improvements in safety and speed for traditional quantification-based repairs, ultimately leading to less overhead cost for the operators. The goals for this research were to develop two types of models with the following goals: 1. Run in real-time on the edge (≥ 12 Hz) 2. Classification: Achieve less than 5% false positive detection 3. Classification: Achieve ≥ 95% methane plume detection rate 4. Regression: achieve ≤ 10 standard cubic feet per hour (scfh) prediction > 70% of the time In order to achieve these results, multiple infrared (IR) and other sensors were investigated in tandem with the midwave IR (MWIR) OGI to provide additional information to train the underlying models. Information on atmospheric conditions including humidity, temperature, pressure, and solar radiation was provided by a weather station. Several machine learning and deep learning architectures and methods, including looking at quantized classification networks and regressions networks, were explored. As further data was collected, curated, and labeled, it allowed for more refined regressive networks to be adequately trained, leading to better insight into the true flow rates being observed. An important valuable deliverable of this research effort was the development of an advanced network which underwent multiple iterations capable of giving a continuous output. The current network has a predicted mean average percentage error (MAPE) of 12.3% just outside our target goal of 10.00%, but an accuracy of 97.78% at ±50 scfh, well within the overall goal for the Department of Energy (DOE) program. Upon closer inspection, it was observed that more than 10% of datapoints contributing to the MAPE predictions were the result of low flow rate predictions and are beyond the sensitivity of instrument measurement as a result of normal operational variation and noise.

03 NATURAL GAS↗

acdc – Automated Contamination Detection and Confidence estimation for single-cell genome data

Abstract Background A major obstacle in single-cell sequencing is sample contamination with foreign DNA. To guarantee clean genome assemblies and to prevent the introduction of contamination into public databases, considerable quality control efforts are put into post-sequencing analysis. Contamination screening generally relies on reference-based methods such as database alignment or marker gene search, which limits the set of detectable contaminants to organisms with closely related reference species. As genomic coverage in the tree of life is highly fragmented, there is an urgent need for a reference-free methodology for contaminant identification in sequence data. Results We present acdc, a tool specifically developed to aid the quality control process of genomic sequence data. By combining supervised and unsupervised methods, it reliably detects both known and de novo contaminants. First, 16S rRNA gene prediction and the inclusion of ultrafast exact alignment techniques allow sequence classification using existing knowledge from databases. Second, reference-free inspection is enabled by the use of state-of-the-art machine learning techniques that include fast, non-linear dimensionality reduction of oligonucleotide signatures and subsequent clustering algorithms that automatically estimate the number of clusters. The latter also enables the removal of any contaminant, yielding a clean sample. Furthermore, given the data complexity and the ill-posedness of clustering, acdc employs bootstrapping techniques to provide statistically profound confidence values. Tested on a large number of samples from diverse sequencing projects, our software is able to quickly and accurately identify contamination. Results are displayed in an interactive user interface. Acdc can be run from the web as well as a dedicated command line application, which allows easy integration into large sequencing project analysis workflows. Conclusions Acdc can reliably detect contamination in single-cell genome data. In addition to database-driven detection, it complements existing tools by its unsupervised techniques, which allow for the detection of de novo contaminants. Our contribution has the potential to drastically reduce the amount of resources put into these processes, particularly in the context of limited availability of reference species. As single-cell genome data continues to grow rapidly, acdc adds to the toolkit of crucial quality assurance tools.

59 BASIC BIOLOGICAL SCIENCES↗

Advanced Concept Studies for Supersonic Commercial Transports Entering Service in the 2018-2020 Period Phase 2

Lockheed Martin Aeronautics Company (LM), working in conjunction with General Electric Global Research (GE GR) and Stanford University, executed a 19 month program responsive to the NASA sponsored "N+2 Supersonic Validation: Advanced Concept Studies for Supersonic Commercial Transports Entering Service in the 2018-2020 Period" contract. The key technical objective of this effort was to validate integrated airframe and propulsion technologies and design methodologies necessary to realize a supersonic vehicle capable of meeting the N+2 environmental and performance goals. The N+2 program is aligned with NASA's Supersonic Project and is focused on providing system level solutions capable of overcoming the efficiency, environmental, and performance barriers to practical supersonic flight. The N+2 environmental and performance goals are outlined in the technical paper, AIAA-2014-2138 (Ref. 1) along with the validated N+2 Phase 2 results. Our Phase 2 efforts built upon our Phase 1 studies (Ref. 2) and successfully demonstrated the ability to design and test realistic configurations capable of shaped sonic booms over the width of the sonic boom carpet. Developing a shaped boom configuration capable of meeting the N+2 shaped boom targets is a key goal for the N+2 program. During the LM Phase 1 effort, LM successfully designed and tested a shaped boom trijet configuration (1021) capable of achieving 85 PLdB under track (forward and aft shock) and up to 28 deg off-track at Mach 1.6. In Phase 2 we developed a refined configuration (1044-2) that extended the under 85 PLdB sonic boom level over the entire carpet of 52 deg off-track at a cruise Mach number of 1.7. Further, the loudness level of the configuration throughout operational conditions calculates to an average of 79 PLdB. These calculations rely on propagation employing Burger's (sBOOM) rounding methodology, and there are indications that the configuration average loudness would actually be 75 PLdB. We also added significant fidelity to the design of the configuration in this phase by performing a low speed wind tunnel test at our LTWT facility in Palmdale, by more complete modelling of propulsion effects in our sonic boom analysis, and by refining our configuration packaging and performance assessments. Working with General Electric, LM performed an assessment of the impact of inlet and nozzle effects on the sonic boom signature of the LM N+2 configurations. Our results indicate that inlet/exhaust streamtube boundary conditions are adequate for conceptual design studies, but realistic propulsion modeling at similar stream-tube conditions does have a small but measurable impact on the sonic boom signature. Previous supersonic transport studies have identified aeroelastic effects as one of the major challenges associated with the long, slender vehicles particularly common with shaped boom aircraft (Ref. 3). Under the Phase 2 effort, we have developed a detailed structural analysis model to evaluate the impact of flexibility and structural considerations on the feasibility of future quiet supersonic transports. We looked in particular at dynamic structural modes and flutter as a failure that must be avoided. We found that for our N+2 design in particular, adequate flutter margin existed. Our flutter margin is large enough to cover uncertainties like large increases in engine weight and the margin is relatively easy to increase with additional stiffening mass. The lack of major aeroelastic problems probably derives somewhat from an early design bias. While shaped boom aircraft require long length, they are not required to be thin. We intentionally developed our structural depths to avoid major flexibility problems. So at the end of Phase 2, we have validated that aeroelastic problems are not necessarily endemic to shaped boom designs. Experimental validation of sonic boom design and analysis techniques was the primary objective of the N+2 Supersonic Validations contract; and in this Phase, LM participated in four high speed wind tunnel tests. The first so-called Parametric Test in the Ames 9x7 tunnel did an exhaustive look at variation effects of the parameters: humidity, total pressure, sample time, spatial averaging distance and number of measurement locations, and more. From the results we learned to obtain data faster and more accurately, and made test condition tolerances easy to meet (eliminating earlier 60 percent wasted time when condition tolerances could not be held). The next two tests used different tunnels. The Ames 11 ft tunnel was used to test lower Mach numbers of 1.2 and 1.4. There were several difficulties using this tunnel for the first time for sonic boom including having to shift the measurement Mach numbers to 1.15 and 1.3 to avoid flow problems. It is believed that the 11 ft could be used successfully to measure sonic boom but there are likely to be a number of test condition restrictions. The Glenn 8x6 ft tunnel was used next and the tunnel has a number of desirable features for sonic boom measurement. While the Ames 9x7 can only test Mach 1.55 to 2.55 and the 11 ft can only test Mach 1.3 and lower, the Glenn 8x6 can test continuously from Mach 0.3 to 2.0. Unfortunately test measurement accuracy was compromised by a reference pressure drift. Post-test analysis revealed that the drift occurred when Mach number drifted slightly. Test measurements indicated that if Mach number drift is eliminated, results from the 8x6 would be more accurate, especially at longer distances, than results from the 9x7. The fourth test in the 9x7, called LM4, used everything we learned to comprehensively and accurately measure our new 1044-02 configuration with a full-carpet shaped signature design. Productivity was 8 times greater than our Phase 1 LM3 test. Measurement accuracy and repeatability was excellent out to 42 in. However, measurements at greater distances require the rail in the aft position and become substantially less accurate. Further signature processing or measurement improvements are needed for beyond near-field signature validation.

Morgenstern, John↗

A Tale of Two Chambers: Iterative Approaches and Lessons Learned from Life Support Systems Testing in Altitude Chambers

The drive for the journey to Mars is in a higher gear than ever before. We are developing new spacecraft and life support systems to take humans to the Red Planet. The journey that development hardware takes before its final incarnation in a fully integrated spacecraft can take years, as is the case for the Orion environmental control and life support system (ECLSS). Through the Pressure Integrated Suit Test (PIST) series, NASA personnel at Johnson Space Center have been characterizing the behavior of a closed loop ECLSS in the event of cabin depressurization. This kind of testing - one of the most hazardous activities performed at JSC - requires an iterative approach, increasing in complexity and hazards). The PIST series, conducted in the Crew and Thermal Systems Division (CTSD) 11-ft Chamber, started with unmanned test precursors before moving to a human-in-the-loop phase, and continues to evolve with the eventual goal of a qualification test for the final system that will be installed on Orion. Meanwhile, the Human Exploration Spacecraft Testbed for Integration and Advancement (HESTIA) program is an effort to research and develop technologies that will work in concert to support habitation on Mars. September 2015 marked the first unmanned HESTIA test, with the goal of characterizing how ECLSS technologies work together in a closed environment. HESTIA will culminate in crewed testing, but it can benefit from the lessons learned from another test that is farther ahead in its development and life cycle. Discussing PIST and HESTIA, this paper illustrates how we approach testing, the kind of information that facility teams need to ensure efficient collaborations and successful testing, and how we can apply what we learn to execute future tests.

Callini, Gianluca↗

Flexible, cluster-based analysis of the electronic medical record of sepsis with composite mixture models

The widespread adoption of electronic medical records (EMRs) in healthcare has provided vast new amounts of data for statistical machine learning researchers in their efforts to model and predict patient health status, potentially enabling novel advances in treatment. In the case of sepsis, a debilitating, dysregulated host response to infection, extracting subtle, uncataloged clinical phenotypes from the EMR with statistical machine learning methods has the potential to impact patient diagnosis and treatment early in the course of their hospitalization. However, there are significant barriers that must be overcome to extract these insights from EMR data. First, EMR datasets consist of both static and dynamic observations of discrete and continuous-valued variables, many of which may be missing, precluding the application of standard multivariate analysis techniques. Second, clinical populations observed via EMRs and relevant to the study and management of conditions like sepsis are often heterogeneous; properly accounting for this heterogeneity is critical. Here, we describe an unsupervised, probabilistic framework called a composite mixture model that can simultaneously accommodate the wide variety of observations frequently observed in EMR datasets, characterize heterogeneous clinical populations, and handle missing observations. In conclusion, we demonstrate the efficacy of our approach on a large-scale sepsis cohort, developing novel techniques built on our model-based clusters to track patient mortality risk over time and identify physiological trends and distinct subgroups of the dataset associated with elevated risk of mortality during hospitalization.

59 BASIC BIOLOGICAL SCIENCES↗

An implantable biomechanical energy harvester for animal monitoring devices

Insufficient service life and the resulting need for battery replacements have been a great challenge for implantable electronic devices. This is particularly true for animal tracking applications, because recapturing animals is often unlikely once they are released to the wild. To tackle this problem, we developed a biomechanical energy harvester that uses a Macro Fiber Composite™ (MFC) piezoelectric beam to harvest the mechanical energy from animals’ body bending movements as the power source for implantable and wearable devices. Prototypes of an underwater acoustic transmitter using this technology were subdermally implanted into juvenile white sturgeon and their energy harvesting performance was evaluated through the device’s transmissions. Additionally, the fish successfully recovered from the implantation surgery and freely swam inside a tank. The transmitter prototypes in the fish continually transmitted signals for a period up to 5 weeks. A benchtop test setup was also created to emulate the fish’s body bending, estimate the device’s energy harvesting performance in the live fish, and perform accelerated fatigue testing of the energy harvester by applying test parameters learned from a video study of the fish’s body movement and behavior characteristics. The gradual depolarization of the piezoelectric ceramic material in the MFC under cyclic mechanical loading was the main limiting factor for the life span of the energy harvester. Pathways for improvement are proposed to achieve long-term efficacy of powering implantable and wearable electronic devices.

42 ENGINEERING↗

Constructing a High‐Resolution Aftershock Catalog for the 2017 Mw 8.2 Tehuantepec Earthquake Sequence Using a Machine Learning–Based Workflow

The 8 September 2017 Mw 8.2 Tehuantepec earthquake was the largest instrumentally recorded normal‐faulting earthquake in Mexico. The mainshock occurred offshore within the Tehuantepec seismic gap, generating >30,000 aftershocks in the following year. We applied an open‐source, machine learning (ML)–assisted workflow to construct a high‐resolution aftershock catalog using data from temporary and permanent seismic networks in southern Mexico. The workflow integrates PhaseNet for phase detection; GaMMA for phase association; and VELEST, HypoInverse, and HypoDD for velocity modeling and relocation. We processed seven months of continuous waveform data from 29 broadband stations, including a temporary rapid‐response deployment that improved station coverage of the offshore rupture zone. To evaluate performance, we compared our results against analyst‐reviewed picks and event locations from the Servicio Sismológico Nacional catalog. The resulting catalog contains 11,374 relocated earthquakes and represents the most comprehensive published dataset for this sequence, incorporating the first full use of the temporary network. Relocated hypocenters show improved depth control and align well with the Slab2.0 subduction geometry, revealing clearer separation between offshore slab events and onshore crustal seismicity. This study demonstrates that combining ML‐based detection with established methods provides a scalable and reproducible approach for constructing high‐quality earthquake catalogs in tectonically complex environments and offers practical guidance for adapting similar workflows to other earthquake sequences.

Garcia, Marc [The University of Texas at El Paso, ↗

Interior Athabascan Energy Capacity Building Network (Final Report)

Tanana Chiefs Conference (TCC) is the traditional tribal consortium for the 37 federally recognized Tribes and villages of Interior Alaska. TCC is based on a belief in tribal self-determination and the need for regional Native unity. TCC is a nonprofit organization that works toward meeting the needs and challenges for more than 10,000 Alaska Natives (mostly Alaskan Athabaskans) in Interior Alaska. With support from DOE-OIE’s Inter-Tribal Technical Assistance grant, TCC’s goal was to expand and build on the solid foundation of work that the existing TCC Energy Program established through the creation of the Interior Athabascan Energy Network (IAEN). More specifically, the IAEN was designed to create a network of village-based energy champions across the interior and hold quarterly teleconferences and annual inperson meetings. Among the greatest impacts of the IAEN were the creation and support of realistic, communitybased energy projects that produced tangible results such as fuel savings and skills and capacity development among local staff as well as establishment of an energy team that included TCC staff, Tribal participants, technical consultants, and support agencies. The DOE-TA funding allowed us to provide technical assistance to our Tribal communities to develop energy projects, and to foster a network consisting of Tribal community members and energy experts that has strengthened the ability in the region to troubleshoot energy issues, develop meaningful projects, envision a clean energy future, and to share project successes and challenges. The funding from DOE was a catalyst for growth, has created opportunities for learning and knowledge sharing, and has led to the development of cost saving energy projects. Most importantly, the DOE project allowed energy champions around the region to share their knowledge, experiences, and expertise to create place-based solutions and has forged inter-Tribal relationships that will continue to be an asset for years to come. Through information exchange that occurred at out IAEN annual meetings and other events, this technical assistance grant award has spurred a spinoff project that is focused on establishing a collaborative of independent utilities that work together to solve energy problems and share services to operate and maintain their electric utilities. We have indentified at least 10 and possibly 12 communities that are interested in collaborating in managing their utilities to reduce costs and increase reliability. As well, by identifying and supporting Community Energy Champions in the Interior villages and adding capacity to the TCC Energy Program, we were able to create actionable community energy plans, develop specific projects, and increase local capacity, skills, and literacy around clean energy initiatives.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗