Search NASA⌕ Search

SEARCH · Search NASA

Results for “feature engineering”

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 109 records · Page 6

Light Duty Engine Performance Characteristics with Dimethyl Ether and Propane

Here, this paper explores the performance characteristics of a compression ignition HYUNDAI 2.2L engine operating with Dimethyl Ether (DME). Test are carried out at three operating conditions that weigh heavily in the FTP75 certification cycle (1000rpm-12Nm, 1500rpm-50Nm, 2000rpm-100Nm). The engine features a high-pressure common rail fuel injection system designed to operate with liquified gases. The main component of the fuel system is a high-pressure pump that incorporates an electronic inlet metering valve commanded on a crank-angle base to control the rail pressure. The pump, which requires no pressure regulator, provides the flow needed to the injectors without flow returning to the inlet. This novel fueling system is leveraged in tests that are conducted to examine the impact of EGR, combustion phasing, injection pressure on efficiency and emissions. In addition, the impact of introducing 15% Propane by mass is examined. During the tests, the engine ECU is aided by an Engine Controller High Speed Oversight unit (ECHO) to provide combustion phasing control, improved cylinder-to-cylinder uniformity, and an effective optimization over the testing effort. The use of DME and Propane allowed for peak thermal efficiency of nearly 43%. These fuels enable significant carbon index (CI) reductions over the baseline Diesel fuel, with indications that 50% reduction in CO 2 over the Diesel engine are possible.

33 ADVANCED PROPULSION SYSTEMS↗

DEEP CELLULAR RECURRENT NEURAL ARCHITECTURE FOR EFFICIENT MULTIDIMENSIONAL TIME-SERIES DATA PROCESSING

Efficient processing of time series data is a fundamental yet challenging problem in pattern recognition. Though recent developments in machine learning and deep learning have enabled remarkable improvements in processing large scale datasets in many application domains, most are designed and regulated to handle inputs that are static in time. Many real-world data, such as in biomedical, surveillance and security, financial, manufacturing and engineering applications, are rarely static in time, and demand models able to recognize patterns in both space and time. Current machine learning (ML) and deep learning (DL) models adapted for time series processing tend to grow in complexity and size to accommodate the additional dimensionality of time. Specifically, the biologically inspired learning based models known as artificial neural networks that have shown extraordinary success in pattern recognition, tend to grow prohibitively large and cumbersome in the presence of large scale multi-dimensional time series biomedical data such as EEG. Consequently, this work aims to develop representative ML and DL models for robust and efficient large scale time series processing. First, we design a novel ML pipeline with efficient feature engineering to process a large scale multi-channel scalp EEG dataset for automated detection of epileptic seizures. With the use of a sophisticated yet computationally efficient time-frequency analysis technique known as harmonic wavelet packet transform and an efficient self-similarity computation based on fractal dimension, we achieve state-of-the-art performance for automated seizure detection in EEG data. Subsequently, we investigate the development of a novel efficient deep recurrent learning model for large scale time series processing. For this, we first study the functionality and training of a biologically inspired neural network architecture known as cellular simultaneous recurrent neural network (CSRN). We obtain a generalization of this network for multiple topological image processing tasks and investigate the learning efficacy of the complex cellular architecture using several state-of-the?art training methods. Finally, we develop a novel deep cellular recurrent neural network (CDRNN) architecture based on the biologically inspired distributed processing used in CSRN for processing time series data. The proposed DCRNN leverages the cellular recurrent architecture to promote extensive weight sharing and efficient, individualized, synchronous processing of multi-source time series data. Experiments on a large scale multi-channel scalp EEG, and a machine fault detection dataset show that the proposed DCRNN offers state-of-the-art recognition performance while using substantially fewer trainable recurrent units.

Vidyaratne, Lasitha S.↗

NRC Consultation and Monitoring at the Savannah River Site: Focusing Reviews of Two Different Disposal Actions - 12181

Section 3116 of the Ronald W. Reagan National Defense Authorization Act for Fiscal Year 2005 (NDAA) requires the U.S. Department of Energy (DOE) to consult with the U.S. Nuclear Regulatory Commission (NRC) for certain non-high level waste determinations. The NDAA also requires NRC to monitor DOE's disposal actions related to those determinations. In Fiscal Year 2011, the NRC staff reviewed DOE performance assessments for tank closure at the F-Tank Farm (FTF) Facility and salt waste disposal at the Saltstone Disposal Facility (SDF) at the Savannah River Site (SRS) as part of consultation and monitoring, respectively. Differences in inventories, waste forms, and key barriers led to different areas of focus in the NRC reviews of these two activities at the SRS. Because of the key role of chemically reducing grouts in both applications, the evaluation of chemical barriers was significant to both reviews. However, radionuclide solubility in precipitated metal oxides is expected to play a significant role in FTF performance whereas release of several key radionuclides from the SDF is controlled by sorption or precipitation within the cementitious wasteform itself. Similarly, both reviews included an evaluation of physical barriers to flow, but differences in the physical configurations of the waste led to differences in the reviews. For example, NRC's review of the FTF focused on the modeled degradation of carbon steel tank liners while the staff's review of the SDF performance included a detailed evaluation of the physical degradation of the saltstone wasteform and infiltration-limiting closure cap. Because of the long time periods considered (i.e., tens of thousands of years), the NRC reviews of both facilities included detailed evaluation of the engineered chemical and physical barriers. The NRC staff reviews of residual waste disposal in the FTF and salt waste disposal in the SDF focused on physical barriers to flow and chemical barriers to radionuclide release from the waste. Because the waste inventory and concentration at both sites is sufficient to generate unacceptable doses to an off-site member of the public or inadvertent intruder in the absence of engineered barriers, the NRC staff review focused on the engineering features DOE plans to put in place to limit radionuclide release. At the FTF, DOE expects that peak doses are delayed beyond a 10,000 year performance period by a combination of (1) the flow-limiting effect of the steel tank liner and (2) chemical conditions created by the stabilizing grout overlying the waste that limit the solubility of key radionuclides for tens of thousands of years. At the SDF, DOE expects that flow will be significantly limited by water shedding along the closure cap lower drainage layer and that radionuclide release will be further limited by radionuclide precipitation or sorption within the high pH, chemically reducing conditions created within the saltstone waste form. Because the performance of both facilities depends on the performance of engineered barriers for thousands of years, the reviews included a detailed evaluation of the expected long-term behavior of these barriers. As previously discussed, NRC staff reviews of DOE waste determinations during consultation are designed to evaluate the three NDAA criteria, whereas the review of an updated PA during monitoring only addresses whether the NRC staff has reasonable assurance that the planned disposal action will meet the performance objectives of 10 CFR Part 61. The NRC staff review of the Waste Determination for the FTF did not include conclusions about whether the planned disposal of residual waste at the FTF would meet the NDAA criteria because of the substantial uncertainties in the degree of waste removal DOE would achieve and other technical uncertainties. The main product of the NRC staff review of the planned FTF disposal action is the recommendation that DOE should conduct waste release experiments to increase support for key modeling assumptions related to: (1) the evolution of pH and Eh in the grouted tank system over time; (2) identification of HRR association with solid phases comprising the residual wastes; and (3) expected solubility of HRRs under a range of environmental or service conditions that the residual wastes in the contaminated zone are expected to be exposed to over time. Implementation of this recommendation is deemed crucial for NRC staff to have reasonable assurance that the performance objectives in 10 CFR Part 61, Subpart C can be met. Given the risk-significance of Tank 18 to the overall PA and the short timeline for closure of this tank, the NRC staff recommended that DOE should initiate discussions with NRC staff regarding implementation of this recommendation for Tank 18 as soon as practical. The NRC staff also recommended that experiments to address this recommendation should be conducted prior to final closure of Tank 18. Results of the Tank 18 residual waste experiments, if conducted, will be evaluated by NRC staff to determine the need for additional data collection, experiments, and modeling for Tank 18, as well as other FTF tanks. Additional information regarding the NRC staff's recommendations in this area, including details on the suggested implementation of other recommendations will be provided in the NRC staff's plan for monitoring the FTF later in FY 2012, after DOE makes a final decision on the waste determination. The NRC staff's review of waste disposal at the SDF is ongoing. When complete, the SDF TER will indicate whether the NRC staff continues to have reasonable assurance that waste disposal at the SDF will meet the performance objectives of 10 CFR Part 61 (NDAA Criterion 3). The TER also will include risk insights that will form the basis of the NRC staff's revised monitoring plan for the SDF. The NRC staff will publish an updated monitoring plan for the SDF later in FY 2012. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Design of the ITER ICRF Antenna

The CYCLE consortium has been designing the ITER ICRF antenna since March 2010, supported by an F4E grant. Following a brief introduction to the consortium, this paper: describes the present status and layout of the design; highlights the key mechanical engineering features; shows the expected impact of cooling and radiation issues on the design and outlines the need for future R and D to support the design process. A key design requirement is the need for the mechanical design and analysis to be consistent with all requirements following from the RF physics and antenna layout optimisation. As such, this paper complements that of Durodie et al.

42 ENGINEERING↗

Predictive analytics of selections of russet potatoes

We explore the application of machine learning algorithms specifically to enhance the selection process of Russet potato (Solanum tuberosum L.) clones in breeding trials by predicting their suitability for advancement. This study addresses the challenge of efficiently identifying high-yield, disease-resistant, and climate-resilient potato varieties that meet processing industry standards. Leveraging manually collected data from trials in the state of Oregon, we investigate the potential of a wide variety of state-of-the-art binary classification models. The dataset includes 1086 clones, with data on 38 attributes recorded for each clone, focusing on yield, size, appearance, and frying characteristics, with several control varieties planted consistently across four Oregon regions from 2013 to 2021. We conduct a comprehensive analysis of the dataset that includes preprocessing, feature engineering, and imputation to address missing values. We focus on several key metrics such as accuracy, F1-score, and Matthews correlation coefficient (MCC) for model evaluation. The top-performing models, namely a feedforward neural network classifier (Neural Net), a histogram-based gradient boosting classifier (HGBC), and a support vector machine classifier (SVM), demonstrate consistent and significant results. To further validate our findings, we conducted a simulation study using the aims, data-generating mechanisms, estimands, methods, and performance measures (ADEMP) framework, simulating different data-generating scenarios to assess model robustness and performance through true positive, true negative, false positive, and false negative distributions, area under the receiver operating characteristic curve (AUC-ROC) and MCC. The simulation results highlight that non-linear models like SVM and HGBC consistently show higher AUC-ROC and MCC than logistic regression, thus outperforming the traditional linear model across various distributions, and emphasizing the importance of model selection and tuning in agricultural trials. Variable selection further enhances model performance and identifies influential features in predicting trial outcomes. The findings emphasize the potential of machine learning in streamlining the selection process for potato varieties, offering benefits such as increased efficiency, substantial cost savings, and judicious resource utilization. Our study contributes insights into precision agriculture and showcases the relevance of advanced technologies for informed decision-making in breeding programs.

60 APPLIED LIFE SCIENCES↗

Beyond Energy Efficiency: A clustering approach to embed demand flexibility into building energy benchmarking

The intermittency of carbon-free renewables and the demand changes associated with the widespread push for electrifying the transportation and building sectors provides an opportunity for buildings to go beyond energy efficiency and push towards providing demand flexibility to the electricity grid. The duality of energy efficiency and demand flexibility is necessary for success in a sustainable and reliable energy transition. Current building energy benchmarking models are limited in their ability to integrate concepts of demand flexibility and/or utilize granular smart meter data. Thus, current benchmarking methods are focused annual energy usage and fail to incorporate how the time of use of energy consumption impacts emissions in a quickly changing energy grid. Without a more comprehensive view of energy usage and associated real-time emissions, current benchmarking methods are unlikely to realize the full decarbonization potential of buildings. New emerging data streams provide an opportunity to develop a new generation of benchmarking energy models that embed dimensions of energy efficiency, grid interactivity, and demand flexibility into their analysis. In this paper, we propose a four-step method for embedding grid interactivity and demand flexibility into building benchmarking models that utilizes emerging building and time-series electricity data streams. We first engineer features to produce a mix-type dataset that encompasses many attributes of grid-interactive and efficient buildings, and then we apply K-medoids using Gower's Distance to produce peer-group clusters. We apply the method to a case study of 306 primary and secondary schools in southern California, USA. The results show that the method effectively clusters buildings by attributes of demand flexibility and energy efficiency. The clustering results reveal patterns in inefficient building operations and demand inflexibility at the building peer group level. In conclusion, the interpretation of clusters can serve as an integrated energy efficiency and demand flexibility benchmarking model and inform performance-specific policy targeting for buildings that go beyond traditional efficiency measures.

24 POWER TRANSMISSION AND DISTRIBUTION↗

SigTime: Learning and Visually Explaining Time Series Signatures

Understanding and distinguishing temporal patterns in time series data is essential for scientific discovery and decision-making. For example, in biomedical research, uncovering meaningful patterns in physiological signals can improve diagnosis, risk assessment, and patient outcomes. However, existing methods for time series pattern discovery face major challenges, including high computational complexity, limited interpretability, and difficulty in capturing meaningful temporal structures. Here, to address these gaps, we introduce a novel learning framework that jointly trains two Transformer models using complementary time series representations: shapelet-based representations to capture localized temporal structures and traditional feature engineering to encode statistical properties. The learned shapelets serve as interpretable signatures that differentiate time series across classification labels. Additionally, we develop a visual analytics system—SigTime—with coordinated views to facilitate exploration of time series signatures from multiple perspectives, aiding in useful insights generation. We quantitatively evaluate our learning framework on eight publicly available datasets and one proprietary clinical dataset. Additionally, we demonstrate the effectiveness of our system through two usage scenarios along with the domain experts: one involving public ECG data and the other focused on preterm labor analysis.

97 MATHEMATICS AND COMPUTING↗

LIFE Materials: Fuel Cycle and Repository Volume 11

The fusion-fission LIFE engine concept provides a path to a sustainable energy future based on safe, carbon-free nuclear power with minimal nuclear waste. The LIFE design ultimately offers many advantages over current and proposed nuclear energy technologies, and could well lead to a true worldwide nuclear energy renaissance. When compared with existing and other proposed future nuclear reactor designs, the LIFE engine exceeds alternatives in the most important measures of proliferation resistance and waste minimization. The engine needs no refueling during its lifetime. It requires no removal of fuel or fissile material generated in the LIFE engine. It leaves no weapons-attractive material at the end of life. Although there is certainly a need for additional work, all indications are that the 'back end' of the fuel cycle does not to raise any 'showstopper' issues for LIFE. Indeed, the LIFE concept has numerous benefits: (1) Per unit of electricity generated, LIFE engines would generate 20-30 times less waste (in terms of mass of heavy metal) requiring disposal in a HLW repository than does the current once-through fuel cycle. (2) Although there may be advanced fuel cycles that can compete with LIFE's low mass flow of heavy metal, all such systems require reprocessing, with attendant proliferation concerns; LIFE engines can do this without enrichment or reprocessing. Moreover, none of the advanced fuel cycles can match the low transuranic content of LIFE waste. (3) The specific thermal power of LIFE waste is initially higher than that of spent LWR fuel. Nevertheless, this higher thermal load can be managed using appropriate engineering features during an interim storage period, and could be accommodated in a Yucca-Mountain-like repository by appropriate 'staging' of the emplacement of waste packages during the operational period of the repository. The planned ventilation rates for Yucca Mountain would be sufficient for LIFE waste to meet the thermal constraints of the repository design. (4) A simple, but arguably conservative, estimate for the dose from a repository containing 63,000 MT of spent LIFE fuel would have similar performance to the currently planned Yucca Mountain Repository. This indicates that a properly designed 'LIFE Repository' would almost certainly meet the proposed Nuclear Regulatory Commission standards for dose to individuals, even though the waste in such a repository would have produced 20-30 times more generated electricity than the reference case for Yucca Mountain. The societal risk/benefit ratio for a LIFE repository would therefore be significantly better than for currently planned repositories for LWR fuel.

42 ENGINEERING↗

Forest aboveground biomass estimation through integration of sentinel-2 and PALSAR-2 time series: assessing models trained on GEDI and field inventory benchmarks

Accurate and spatially explicit forest Aboveground Biomass (AGB) mapping through remote sensing is critical for quantifying terrestrial carbon stocks and informing effective forest management strategies. However, AGB estimation in dense forests with complex terrain remains challenging due to satellite sensor signal saturation problem (saturation issue occurs in high biomass forests), structural complexity, and limited ground truth for calibration. This study presents a novel framework that integrates multi-temporal Sentinel-2 optical imagery, ALOS PALSAR-2 Synthetic Aperture Radar (SAR) data, and topographic variables with explainable Machine Learning to map AGB across mountainous forests within subtropical and temperate oceanic climate zones of Mexico. We evaluate the effects of temporal granularity and sensor synergy by comparing multiple temporal inputs and sensor configurations (Sentinel-2, PALSAR-2, and their fusion), and assess model performance using two reference datasets: NASA GEDI LiDAR-derived biomass and Mexico’s National Forest and Soil Inventory (INFyS). Our results showed that models trained on INFyS consistently outperformed those trained on GEDI, highlighting limitations in GEDI’s reliability in biomass estimates within this study region. Furthermore, the integration of Sentinel-2 and PALSAR-2 provided improved predictions compared to single-sensor models, particularly when combined with temporally explicit yearly statistics. The best-performing model, which was trained on INFyS data, and considered both Sentinel-2 and PALSAR-2 yearly statistics, as well as topographic variables, achieved an R2 of 0.64, RMSE of 51.10 Mg/ha, and relative RMSE (rRMSE) of 58.69%. Explainable ML analysis identified Sentinel-2 spectral indices and topographic features as key predictors, while PALSAR-2 metrics provided complementary information, partially mitigating saturation effects in high-biomass areas. Specifically, integrating both sensors substantially improved AGB estimation in high biomass forest (≥200 Mg/ha), yielding 98% gains over optical-only model, with resulting estimates exceeding GEDI L4B by 29% and ESA-CCI-BIOMASS by 174%. Terrain-stratified analysis indicated close agreement with GEDI in low-slope areas, with increasing divergence as slope steepness increased, while estimates remained consistently higher than ESA-CCI-BIOMASS across all slope classes. The proposed approach advances multi-sensor fusion and temporal feature engineering for AGB mapping using open-access satellite datasets, providing a scalable and reproducible framework for annual biomass monitoring in topographically complex mountainous forests. The resulting 25 m resolution biomass product has the potential to provide spatially detailed information for forest monitoring and may support applications in carbon accounting and forest management.

54 ENVIRONMENTAL SCIENCES↗

Mix and hydrodynamic instabilities on NIF

Several new platforms have been developed to experimentally measure hydrodynamic instabilities in all phases of indirect-drive, inertial confinement fusion implosions on National Ignition Facility. At the ablation front, instability growth of pre-imposed modulations was measured with a face-on, x-ray radiography platform in the linear regime using the Hydrodynamic Growth Radiography (HGR) platform. Modulation growth of "native roughness" modulations and engineering features (fill tubes and capsule support membranes) were measured in conditions relevant to layered DT implosions. A new experimental platform was developed to measure instability growth at the ablator-ice interface. Here in the deceleration phase of implosions, several experimental platforms were developed to measure both low-mode asymmetries and high-mode perturbations near peak compression with x-ray and nuclear techniques. In one innovative technique, the self-emission from the hot spot was enhanced with argon dopant to "self-backlight" the shell in-flight. To stabilize instability growth, new "adiabat-shaping" techniques were developed using the HGR platform and applied in layered DT implosions.

42 ENGINEERING↗

Dimensionally Aligned Signal Projection Algorithms Library

Dimensionally aligned signal projection (DASP) algorithms are used to analyze fast Fourier transforms (FFTs) and generate visualizations that help focus on the harmonics for specific signals. At a high level, these algorithms extract the FFT segments around each harmonic frequency center, and then align them in equally sized arrays ordered by increasing distance from the base frequency. This allows for a focused view of the harmonic frequencies, which, among other use cases, can enable machine learning algorithms to more easily identify salient patterns. This work seeks to provide an effective open-source implementation of the DASP algorithms proposed by Vann et al. (2018) as well as functionality to help explore and test how these algorithms work with an interactive dashboard and signal-generation tool. The DASP library is implemented in Python and contains four types of algorithms for implementing these feature engineering techniques: fixed harmonically aligned signal projection (HASP), decimating HASP, interpolating HASP, and frequency aligned signal projection (FASP). Each algorithm returns a numerical array, which can be visualized as an image. The HASP algorithms are variations of the algorithms originally presented by Vann et al. (2018). For consistency, FASP, which is the terminology used for the short-time Fourier transform (STFT), has been implemented as part of the library to provide a similar interface to the STFT of the raw signal. Additionally, the library contains an algorithm to generate artificial signals with basic customizations such as the base frequency, sample rate, duration, number of harmonics, noise, and number of signals. Finally, the library provides multiple interactive visualizations, each of which is implemented using IPyWidgets and works in a Jupyter environment. A dashboard-style visualization is provided, which contains some common signal-processing visual components (signal, FFT, spectogram) updating in unison with the HASP functions (see Figure 1 below). Separate from the dashboard, an independent visualization is provided for each of the DASP algorithms as well as for the artifical signal generator. These visualizations are included in the library to aid in developing an intuitive understanding how the algorithms are affected by different input signals and parameter selections.

harmonics↗

Transfer Printing Techniques for Materials Assembly and Micro/Nanodevice Fabrication

Transfer printing represents a set of techniques for deterministic assembly of micro-and nanomaterials into spatially organized, functional arrangements with two and three-dimensional layouts. Such processes provide versatile routes not only to test structures and vehicles for scientific studies but also to high-performance, heterogeneously integrated functional systems, including those in flexible electronics, three-dimensional and/or curvilinear optoelectronics, and bio-integrated sensing and therapeutic devices. This article summarizes recent advances in a variety of transfer printing techniques, ranging from the mechanics and materials aspects that govern their operation to engineering features of their use in systems with varying levels of complexity. Here, a concluding section presents perspectives on opportunities for basic and applied research, and on emerging use of these methods in high throughput, industrial-scale manufacturing.

36 MATERIALS SCIENCE↗

Data‐Driven Insights into Rare Earth Mineralization: Machine Learning Applications Using Functional Material Synthesis Data

Understanding rare‐earth element (REE) mineralization mechanisms is essential for developing efficient separation strategies. Although the geochemical pathways that generate REE deposits are qualitatively known, quantitative links between specific conditions and mineralization outcomes remain limited. Herein, the repurpose laboratory REE hydrothermal synthesis data—originally collected for functional‐materials fabrication—as a surrogate for studying mineralization with data‐driven methods. The compiled 1,200+ hydrothermal reaction records and trained three machine‐learning models—K‐nearest neighbors (KNN), random forest (RF), and extreme gradient boosting (XGB)—to predict product elements and phases from precursors, additives, reaction conditions, and engineered features. Validation shows XGB achieves the highest accuracy. Feature importance indicates thermodynamic properties of cations and anions dominate model decisions. Correlations reveal positive relationships among precursor concentration, reaction time, pH, and temperature, consistent with classical crystallization behavior. XGB‐based regressors are built to predict crystallization temperature and pH from precursor/product attributes. Performance is strongest when similar training examples exist, while accuracy declines for underrepresented reactions, notably REE carbonates and heavy‐REE systems. Overall, the study shows that functional‐materials datasets can illuminate REE mineralization and provide priors for exploration and processing. Expanding datasets with less‐studied chemistries and conditions will improve generality and support deposit discovery and more efficient REE recovery.

feature importance analysis↗

In-situ sensor monitoring of multi-class gas porosity formation in laser powder bed fusion using convolutional neural network

In-situ monitoring of defect formation remains a significant challenge in the laser powder bed fusion (LPBF) process. Recent advances have enabled real-time defect detection with machine learning and in-situ sensing technologies; however, most studies focus on binary classification of keyhole pores, limiting nuanced multi-class pore differentiation and formation mechanisms. This work introduces a multi-class pore detection framework (no pore, small pores < 15 µm, and large pores > 15 µm) by leveraging photodiode sensor data alongside high-fidelity synchrotron X-ray imaging. The 15 µm threshold is selected to distinguish between two fundamentally different defect mechanisms, following the physical size-mechanism boundary established by prior high-resolution synchrotron X-ray characterization of Al6061 LPBF. Distinguishing these classes is critical because large keyhole pores are structurally detrimental, whereas small gas pores are often benign, requiring different process control strategies. Thermal emission monitoring data collected simultaneously with high-speed X-ray imaging at the Stanford Synchrotron Radiation Lightsource (SSRL), are correlated with subsurface melt pool dynamics to establish ground truth. Continuous Wavelet Transform (CWT) with optimized parameters converts the photodiode time-series signals into time–frequency images, facilitating feature extraction. Convolutional Neural Networks (CNN) are then applied for real-time multi-class pore classification in an average inference time of 1 ms per signal window. It achieves 79% accuracy and an Area Under the Receiver Operating Characteristic curve (AUC ROC) score of 0.89 with five-fold cross-validation. The results demonstrate that coupling CWT-based feature engineering with CNN architecture enables reliable multi-class pore detection in Al6061 builds using affordable in-situ sensors. This approach advances scalable and affordable quality assurance in additive manufacturing by moving beyond binary defect detection toward more nuanced classification of porosity mechanisms with in-situ sensors and machine learning.

Laser powder bed fusion, Multi-class pores, In-sit↗

Inferring occupant counts from Wi-Fi data in buildings through machine learning

An important approach to curtail building energy consumption is to optimize building control based on occupancy information. Various studies proposed to estimate occupant counts through different approaches and sensors. However, high cost and privacy concerns remain as major barriers, restricting the practice of occupant count detection. In this study, we propose a novel method utilizing data from widely deployed Wi-Fi infrastructure to infer occupant counts through machine learning. Compared with the current indirect measurement methods, our method improves the performance of estimating people count: (1) we avoid privacy concerns by anonymizing and reshuffling the MAC addresses on a daily basis; (2) we adopted a heuristic feature engineer approach to cluster connected devices into different types based on their daily connection duration. We tested the method in an office building located in California. In an area with an average occupancy of 22–27 people and a peak occupancy of 48 - 74 people, the root square mean error on the test set is less than four people. The error is within two people counts for more than 70% of estimations, and less than six counts for more than 90% of estimations, indicating a relatively high accuracy. The major contribution of this study is proposing a novel and accurate approach to detect occupant counts in a non-intrusive way, i.e., utilizing existing Wi-Fi infrastructure in buildings without requiring the installation of extra hardware or sensors. The method we proposed is generic and could be applied to other commercial buildings to infer occupant counts for energy efficient building control.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Reduced-order modeling through machine learning and graph-theoretic approaches for brittle fracture applications

Typically, thousands of computationally expensive micro-scale simulations of brittle crack propagation are needed to upscale lower length scale phenomena to the macro-continuum scale. Running such a large number of crack propagation simulations presents a significant computational challenge, making reduced-order models (ROMs) attractive for this task. Here, the ultimate goal of this research is to develop ROMs that have sufficient accuracy and low computational cost so that these upscaling simulations can be readily performed. However, constructing ROMs for these complex simulations presents its own challenge. Here, we present and compare four different approaches for reduced-order modeling of brittle crack propagation in geomaterials. These methods rely on machine learning (ML) and graph-theoretic algorithms to approximate key aspects of the brittle crack problem. These methods also incorporate different physics-based assumptions in order to reduce the training requirements while maintaining accurate physics as much as possible. Results from the ROMs are directly compared against a high-fidelity model of brittle crack propagation. Further, the strengths and weaknesses of the ROMs are discussed, and we conclude that combining smart physics-informed feature engineering with highly trainable ML models provides the best performance. The ROMs considered here have computational costs that are orders-of-magnitude less than the cost associated with high-fidelity physical models while maintaining good accuracy.

36 MATERIALS SCIENCE↗

Detecting control system misbehavior by fingerprinting programmable logic controller functionality

In recent years, attacks such as the Stuxnet malware have demonstrated that cyberattacks against control systems cause extensive damage. These attacks can result in physical damage to the networked systems under their control. In this paper, we discuss our approach for detecting such attacks by distinguishing between programs running on a programmable logic controller (PLC) without having to monitor communications. Using power signatures generated by an attached, high-frequency power measurement device, we can identify what a PLC is doing and when an attack may have altered what the PLC should be doing. To accomplish this, we generated labeled data for testing our methods and applied feature engineering techniques and machine learning models. The results of this research demonstrate that Random Forests and Convolutional Neural Networks classify programs with up to 98% accuracy for major program differences and 84% accuracy for minor differences. Our results can be used for both online and offline applications.

97 MATHEMATICS AND COMPUTING↗

End-To-End Decentralized Transmission Line Protection in IBR-Dominated Weak Grids Using Interpretable Data-Driven Methods

Traditional transmission line protection relies on predictable synchronous-based fault signatures, which frequently fail under the non-standard, current-limited fault characteristics of Inverter-Based Resources (IBRs). This study investigates how to achieve secure, communication-free fault isolation in IBR-dominated weak grids without relying on opaque, computationally heavy "black-box" machine learning algorithms. To address this, we propose a novel, standalone, and inherently interpretable data-driven protection framework. Unlike centralized methods requiring multi-terminal communication, this decentralized approach relies solely on local measurements using a hierarchical linear-kernel Support Vector Machine (SVM). The methodology decomposes the protection task into four sequential stages that mimic traditional protection elements: fault detection and fault direction identification, fault type classification, zone classification, and location estimation. This multi-stage architecture allows for specialized feature engineering at each stage, combining high computational efficiency with logic traceability. The framework's end-to-end performance was validated via C-code and PSCAD/EMTDC co-simulation, utilizing a real-world utility network and an OEM black-box IBR model. The proposed relay achieves 97.2% overall accuracy and provides a reliable trip decision within a 2.5-cycle window. The results confirm 100% accuracy in fundamental fault detection, reliable zone selectivity across low to moderate fault resistances, and robust security against non-fault transients, proving its immediate viability for integration into commercial numerical relays.

24 POWER TRANSMISSION AND DISTRIBUTION↗