Search NASA⌕ Search

SEARCH · Search NASA

Results for “statistical 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 1,783 records · Page 99

End-to-end online performance data capture and analysis for scientific workflows

With the increased prevalence of employing workflows for scientific computing and a push towards exascale computing, it has become paramount that we are able to analyze characteristics of scientific applications to better understand their impact on the underlying infrastructure and vice-versa. Such analysis can help drive the design, development, and optimization of these next generation systems and solutions. Here, we present the architecture, integrated with existing well-established and newly developed tools, to collect online performance statistics of workflow executions from various, heterogeneous sources and publish them in a distributed database (Elasticsearch). Using this architecture, we are able to correlate online workflow performance data, with data from the underlying infrastructure, and present them in a useful and intuitive way via an online dashboard. We have validated our approach by executing two classes of real-world workflows, both under normal and anomalous conditions. The first is an I/O-intensive genome analysis workflow; the second, a CPU- and memory-intensive material science workflow. Based on the data collected in Elasticsearch, we are able to demonstrate that we can correctly identify anomalies that we injected. The resulting end-to-end data collection of workflow performance data is an important resource of training data for automated machine learning analysis.

97 MATHEMATICS AND COMPUTING↗

Comparing Designed Training Sets to Optimize Multivariate Regression Models for Pr, Nd, and Nitric Acid Using Spectrophotometry

Chemometric regression models were developed for the quantification of praseodymium (Pr, 0–1000 µg/mL), neodymium (Nd, 0–1000 µg/mL), and nitric acid (HNO 3 , 0.1–5 M) using spectrophotometry. Designed calibration sets were composed of 20 samples each: 10 model points and 10 lack-of-fit (LOF) points. The D-optimal designs effectively minimized the number of samples required to build models, and each design resulted in similar prediction performance, suggesting that statistical design of experiments can provide a reliable framework for selecting training set samples in three-variable systems. Partial least squares regression (PLSR) models were validated against a one-factor-at-a-time validation set composed of 125 samples (three variables, five levels). The top PLS-1 models resulted in average percent root mean square error of prediction error values of 3.5%, 1.7%, and 1.2% for Pr(III), Nd(III), and HNO 3 , respectively. Power set augmentations of the model and LOF samples were investigated to optimize the number of training set samples. PLSR models built using just required model points (10) had similar predictive capabilities as models including the LOF points (20) but with fewer samples. The number of validation samples was also varied systematically to learn how many samples are needed to validate regression models. This work addresses long-standing questions in the field of chemometrics to help make this approach amenable to the near-real-time quantification of hazardous species in remote settings.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Statistical analysis of the chemical attribution signatures of 3-methylfentanyl and its methods of production

Chemical attribution of the origin of an illegal drug is a key component of forensic efforts aimed at combating illicit and clandestine manufacture of drugs and pharmaceuticals. The results of these studies yield detailed information on synthesis byproducts, reagents, and precursors that can be used to identify the method of manufacture. Here in the present work, chemical attribution signatures (CAS) associated with the synthesis of the analgesic 3-methylfentanyl, N-(3-methyl-1-phenethylpiperidin-4-yl)-N-phenylpropanamide, were investigated. Eighteen crude samples from six synthesis methods were generated, the analysis of which was used to identify signatures (i.e. chemical compounds) that were important in the discrimination of synthetic route. These methods were carefully selected to minimize the use of scheduled precursors, complicated laboratory equipment, number of steps, and extreme reaction conditions. Using gas and liquid chromatographies combined with time-of-flight mass spectrometry (GC-QTOF and LC-QTOF) over 160 distinct species were monitored. Lastly, analysis of this combined data set was performed using modern machine learning techniques capable of reducing the size of the data set, prioritizing key chemical attribution signatures, and identifying the method of production for blindly synthesized 3-methylfentanyl materials.

3-methylfentanyl↗

Tightening the thermodynamic uncertainty relations with null-entropy events: What we learn when nothing happens

Fluctuation theorems establish that thermodynamic processes at the microscale can occasionally result in negative entropy production. At the microscale, another distinct possibility becomes more likely: processes where no entropy is produced overall. In this work, we explore the constraints imposed by such null-entropy events on the fluctuations of thermodynamic currents. By incorporating the probability of null-entropy events, we obtain tighter bounds on finite-time thermodynamic uncertainty relations derived from fluctuation theorems. We validate this framework using an example of a qudit SWAP engine.

FOS: Physical sciences↗

Photometric redshift estimation with convolutional neural networks and galaxy images: Case study of resolving biases in data-driven methods

Deep-learning models have been increasingly exploited in astrophysical studies, but these data-driven algorithms are prone to producing biased outputs that are detrimental for subsequent analyses. In this work, we investigate two main forms of biases: class-dependent residuals, and mode collapse. We do this in a case study, in which we estimate photometric redshift as a classification problem using convolutional neural networks (CNNs) trained with galaxy images and associated spectroscopic redshifts. We focus on point estimates and propose a set of consecutive steps for resolving the two biases based on CNN models, involving representation learning with multichannel outputs, balancing the training data, and leveraging soft labels. The residuals can be viewed as a function of spectroscopic redshift or photometric redshift, and the biases with respect to these two definitions are incompatible and should be treated individually. We suggest that a prerequisite for resolving biases in photometric space is resolving biases in spectroscopic space. Experiments show that our methods can better control biases than benchmark methods, and they are robust in various implementing and training conditions with high-quality data. Our methods hold promises for future cosmological surveys that require a good constraint of biases, and they may be applied to regression problems and other studies that make use of data-driven models. Nonetheless, the bias-variance tradeoff and the requirement of sufficient statistics suggest that we need better methods and optimized data usage strategies.

79 ASTRONOMY AND ASTROPHYSICS↗

Quantitative and interpretable order parameters for phase transitions from persistent homology

Here, we apply modern methods in computational topology to the task of discovering and characterizing phase transitions. As illustrations, we apply our method to four two-dimensional lattice spin models: the Ising, square ice, XY, and fully frustrated XY models. In particular, we use persistent homology, which computes the births and deaths of individual topological features as a coarse-graining scale or sublevel threshold is increased, to summarize multiscale and high-point correlations in a spin configuration. We employ vector representations of this information called persistence images to formulate and perform the statistical task of distinguishing phases. For the models we consider, a simple logistic regression on these images is sufficient to identify the phase transition. Interpretable order parameters are then read from the weights of the regression. This method suffices to identify magnetization, frustration, and vortex-antivortex structure as relevant features for phase transitions in our models. We also define “persistence” critical exponents and study how they are related to those critical exponents usually considered.

36 MATERIALS SCIENCE↗

Automated Quantitative Analysis of Extended Irradiation Defects - Dislocations, Voids and Precipitates in Neutron Irradiated HT-9 Steel

Structural alloys that make up the core components of nuclear reactors are exposed to radiation that can substantially alter the microstructure and mechanical properties of the material. Irradiation can cause the formation of extended defects including dislocation line and loops, voids and various types of precipitates. To understand the irradiation effects, it is critical to characterize these defects individually and collectively to quantify how they evolve throughout irradiation stages. Transmission electron microscopy (TEM) is a standard tool for providing extensive size, number density and spatial distribution of extended irradiation defects, however, manual measurement of defect size and number density requires time-intensive effort to analyze enough defects that provide a statistically meaningful result and are prone to errors of subjectivity. In this work, we developed a set of MATLAB algorithms offering robust and automated identification and quantification of extended irradiation defects like dislocation lines, voids and precipitates commonly observed in neutron irradiated cladding alloys [1]. Using a neutron-irradiated HT-9 martensitic alloy irradiated at 412°C to 111.8 dpa, we demonstrate the capability of these algorithms for quantitative defect analysis. Figure 1 presents defect images of the HT-9 using an advanced diffraction contrast imaging scanning transmission electron microscopy (DCI STEM) that suppresses bend contours and provides better defect contrast compared to conventional TEM [2]. The first step of defect analysis is defect identification, a process that can be digitally achieved by contrast thresholding (for simple defect contrast) and by machine learning computer vision for defect semantic segmentation [3]. Here, labeled binary defect images (Figures 2a and 2c) were used. For dislocation density quantification, we employed an optimized circular-grid intersection method first introduced by Dr. John Steeds [4]. This method is shown in Figures 2a and 2b, where the number of intersections of dislocation lines with circular traces was measured. This produces an unbiased sampling of dislocation lines, regardless of their orientation. This process was optimized with automated digital intersection number and grid length registration. Interestingly, it was found that a relatively large number of circles in a set circular gird is was needed to produce an invariant result (Figure 2b), with a low systematic measurement error. This demonstrates the main advantage of computer-based defect analysis, in performing repetitive operations quickly with a reduced systematic error. For the quantitative analysis of voids and precipitates a challenge encountered with automated defect analysis is feature overlap, particularly when the density of voids/precipitates is high. To mitigate this problem, as shown in Figure 2d, we modified the popular circular Hough transformation and treated ‘circular shape’ as the particle descriptor for voids/precipitates, efficiently resolving the overlapping particles. In summary, this automated digital quantification together with machine learning enabled defect identification [3] opens new opportunities for standardizing the characterization workflow for quantitative analysis of extended irradiation defects.

Sainju, Rajat↗

Cable Damage Detection System and Algorithms Using Time Domain Reflectometry

This report describes the hardware system and the set of algorithms we have developed for detecting damage in cables for the Advanced Development and Process Technologies (ADAPT) Program. This program is part of the W80 Life Extension Program (LEP). The system could be generalized for application to other systems in the future. Critical cables can undergo various types of damage (e.g. short circuits, open circuits, punctures, compression) that manifest as changes in the dielectric/impedance properties of the cables. For our specific problem, only one end of the cable is accessible, and no exemplars of actual damage are available. This work addresses the detection of dielectric/impedance anomalies in transient time domain reflectometry (TDR) measurements on the cables. The approach is to interrogate the cable using time domain reflectometry (TDR) techniques, in which a known pulse is inserted into the cable, and reflections from the cable are measured. The key operating principle is that any important cable damage will manifest itself as an electrical impedance discontinuity that can be measured in the TDR response signal. Machine learning classification algorithms are effectively eliminated from consideration, because only a small number of cables is available for testing; so a sufficient sample size is not attainable. Nonetheless, a key requirement is to achieve very high probability of detection and very low probability of false alarm. The approach is to compare TDR signals from possibly damaged cables to signals or an empirical model derived from reference cables that are known to be undamaged. This requires that the TDR signals are reasonably repeatable from test to test on the same cable, and from cable to cable. Empirical studies show that the repeatability issue is the 'long pole in the tent' for damage detection, because it is has been difficult to achieve reasonable repeatability. This one factor dominated the project. The two-step model-based approach is summarized as follows: Step 1, Cable Modeling: Given input-output TDR signals s(n) and x{sub U}(n) for a cable known to be free of damage, system identification algorithms are used to compute a dynamic prediction-error cable model that has output {cflx x}{sub U}(n). The model is declared valid when the innovations e{sub U}(n) = x{sub U}(n) {cflx x}{sub U}(n) satisfy a statistical zero-mean whiteness test. This validated model output is then used as a known reference to which other cables can be compared. Step 2, Cable Testing: The TDR output signal x{sub D}(n) from a cable under test is compared with the model output {cflx x}{sub U}(n) by computing the innovations e{sub D}(n) = x{sub D}(n) {cflx x}{sub U}(n). The innovations are tested using a short-term whiteness test statistic, which employs a statistical confidence interval. If the cable passes the test, this implies that the model is valid and the cable is declared undamaged. If the cable fails the test, this indicates a model mismatch, which means that the cable's dielectric properties have changed; and this implies that the cable is damaged. The test threshold is adjusted to maximize probability of detection and minimize probability of false alarm according to an empirically determined receiver operating characteristic (ROC) curve. An associated confidence interval on the probability of correct classification is also provided. The effectiveness of the algorithm is demonstrated using measured TDR signals for undamaged and damaged cables. Experimental and algorithmic methods for coping with repeatability issues are presented. The model-based damage detection algorithms have been shown to perform well for some representative examples of real TDR signals acquired using the two-dimensional (2D) mockup fixture. If the damage causes a short circuit, then damage detection performance is generally good to excellent. Examples include the cases demonstrated in this report for cuts and pinholes. If the damage does not cause a short circuit, then damage detection performance is generally poor to fair. Examples include the cases demonstrated in this report for crushes and partial cuts.

42 ENGINEERING↗

Target Detection via Cognitive Radars Using Change-Point Detection, Learning, and Adaptation

Many radar detection algorithms that assume a stationary environment (clutter) have been proposed and analyzed over the years. However, in practice, changes in the nonstationary environment can perturb the parameters of the clutter distribution, or even alter the clutter distribution family, which can greatly deteriorate the target detection capability. To avoid such potential performance degradation, cognitive radar systems are envisioned which are required to rapidly realize the nonstationarity, accurately learn the new characteristics of the environments, and adaptively update the detector. In this paper, aiming to develop a fully cognitive radar for target detection in nonstationary environments, we propose a unifying framework that integrates (i) change-point detection of clutter distributions by using a data-driven cumulative sum (CUSUM) algorithm and its extended version, (ii) learning/identification of clutter distribution by applying sparse theory and kernel density estimation methods, and (iii) adaptive target detection by automatically modifying the likelihood-ratio test and corresponding detection threshold. Further, with extensive numerical examples, we demonstrate the achieved improvements in detection performance due to the proposed framework in comparison to a nonadaptive case, an adaptive matched filter (AMF) method, and the clairvoyant case. Herein, we also use Wilcoxon rank-sum tests to evaluate the statistical significance of the performance improvements

42 ENGINEERING↗

Statistical mentoring at early training and career stages

At Los Alamos National Laboratory (LANL), statistical scientists develop solutions for a variety of national security challenges through scientific excellence, typically as members of interdisciplinary teams. At LANL, mentoring is actively encouraged and practiced to develop statistical skills and positive career-building behaviors. Mentoring activities targeted at different career phases from student to junior staff are an important catalyst for both short and long term career development. This article discusses mentoring strategies for undergraduate and graduate students through internships as well as for postdoctoral research associates and junior staff. Topics addressed include project selection, progress, and outcome; intellectual and social activities that complement the student internship experience; key skills/knowledge not typically obtained in academic training; and the impact of such internships on students’ careers. Experiences and strategies from a number of successful mentorships are presented. Feedback from former mentees obtained via a questionnaire is incorporated. As a result, these responses address some of the benefits the respondents received from mentoring, helpful contributions and advice from their mentors, key skills learned, and how mentoring impacted their later careers.

97 MATHEMATICS AND COMPUTING↗

Selection Algorithm Improvement for MicroBooNE

Data selection is an extremely important part of data analysis for any experiment. Finding a physics result is often the result of sifting through a massive amount of data, keeping data that we believe to be signal and throwing out data we do not. This process is called data selection. Creating a selection algorithm is an intensive process that must balance keeping enough data to have statistics and maximizing the signal purity of that data. In this study, we used three different reconstruction tools, Pandora, WireCell, and LANTERN, for the MicroBooNE experiment in conjunction to improve the selection algorithm for analysis. For the case of this study, we look into the charged current N proton 0 pions (CCNp0$\pi$) interaction channel. This is the dominant channel for the Short Baseline Neutrino (SBN) program and is expected to be a large contributor to the Deep Underground Neutrino Experiment (DUNE). We first investigated each of the three tools to find out more about their strengths and weaknesses as reconstructions. We then put together a direct comparison of the three methods to find which method or combination of methods would return the best result for us. While the study is ongoing, we have learned a lot about data selection for the experiment and the differences between the reconstruction tools.

Dillon, Brayden [Michigan State U.]↗

A differentiable simulation package for performing inference of synchrotron-radiation-based diagnostics

The direction of particle accelerator development is ever-increasing beam quality, currents and repetition rates. This poses a challenge to traditional diagnostics that directly intercept the beam due to the mutual destruction of both the beam and the diagnostic. An alternative approach is to infer beam parameters non-invasively from the synchrotron radiation emitted in bending magnets. However, inferring the beam distribution from a measured radiation pattern is a complex and computationally expensive task. To address this challenge we present SYRIPY ( SYnchrotron Radiation In PYthon ), a software package intended as a tool for performing inference of synchrotron-radiation-based diagnostics. SYRIPY has been developed using PyTorch , which makes it both differentiable and able to leverage the high performance of GPUs, two vital characteristics for performing statistical inference. The package consists of three modules: a particle tracker, Lienard–Wiechert solver and Fourier optics propagator, allowing start-to-end simulation of synchrotron radiation detection to be carried out. SYRIPY has been benchmarked against SRW , the prevalent numerical package in the field, showing good agreement and up to a 50× speed improvement. Finally, we have demonstrated how SYRIPY can be used to perform Bayesian inference of beam parameters using stochastic variational inference.

43 PARTICLE ACCELERATORS↗

Outlier classification using autoencoders: application for fluctuation driven flows in fusion plasmas

Understanding the statistics of fluctuation driven flows in the boundary layer of magnetically confined plasmas is desired to accurately model the lifetime of the vacuum vessel components. Mirror Langmuir probes (MLPs) are a novel diagnostic that uniquely allow us to sample the plasma parameters on a time scale shorter than the characteristic time scale of their fluctuations. Sudden large-amplitude fluctuations in the plasma degrade the precision and accuracy of the plasma parameters reported by MLPs for cases in which the probe bias range is of insufficient amplitude. While some data samples can readily be classified as valid and invalid, we find that such a classification may be ambiguous for up to 40% of data sampled for the plasma parameters and bias voltages considered in this study. In this contribution, we employ an autoencoder (AE) to learn a low-dimensional representation of valid data samples. By definition, the coordinates in this space are the features that mostly characterize valid data. Ambiguous data samples are classified in this space using standard classifiers for vectorial data. In this way, we avoid defining complicated threshold rules to identify outliers, which require strong assumptions and introduce biases in the analysis. By removing the outliers that are identified in the latent low-dimensional space of the AE, we find that the average conductive and convective radial heat fluxes are between approximately 5% and 15% lower as when removing outliers identified by threshold values. For contributions to the radial heat flux due to triple correlations, the difference is up to 40%.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Space Shuttle Day-of-Launch Trajectory Design Operations

A top priority of any launch vehicle is to insert as much mass into the desired orbit as possible. This requirement must be traded against vehicle capability in terms of dynamic control, thermal constraints, and structural margins. The vehicle is certified to specific structural limits which will yield certain performance characteristics of mass to orbit. Some limits cannot be certified generically and must be checked with each mission design. The most sensitive limits require an assessment on the day-of-launch. To further minimize vehicle loads while maximizing vehicle performance, a day-of-launch trajectory can be designed. This design is optimized according to that day s wind and atmospheric conditions, which increase the probability of launch. The day-of-launch trajectory design and verification process is critical to the vehicle s safety. The Day-Of-Launch I-Load Update (DOLILU) is the process by which the National Aeronautics and Space Administration's (NASA) Space Shuttle Program tailors the vehicle steering commands to fit that day s environmental conditions and then rigorously verifies the integrated vehicle trajectory s loads, controls, and performance. This process has been successfully used for almost twenty years and shares many of the same elements with other launch vehicles that execute a day-of-launch trajectory design or day-of-launch trajectory verification. Weather balloon data is gathered at the launch site and transmitted to the Johnson Space Center s Mission Control. The vehicle s first stage trajectory is then adjusted to the measured wind and atmosphere data. The resultant trajectory must satisfy loads and controls constraints. Additionally, these assessments statistically protect for non-observed dispersions. One such dispersion is the change in the wind from the last measured balloon to launch time. This process is started in the hours before launch and is repeated several times as the launch count proceeds. Should the trajectory design not meet all constraint criteria, Shuttle would be No-Go for launch. This Shuttle methodology is very similar to other unmanned launch vehicles. By extension, this method would likely be employed for any future NASA launch vehicle. This paper will review the Shuttle s day-of-launch trajectory optimization and verification operations as an example of a more generic application of day-of-launch design and validation. With Shuttle s retirement, it is fitting to document the current state of this critical process and capture lessons learned to benefit current and future launch vehicle endeavors.

Harrington, Brian E.↗

How much can we learn from a merging cold front cluster? Insights from X-ray temperature and radio maps of A3667

The galaxy cluster A3667 is an ideal laboratory to study the plasma processes in the intracluster medium. High-resolution Chandra X-ray observations show a cold front in A3667. At radio wavelengths, A3667 reveals a double radio-relic feature in the outskirts of the cluster. These suggest multiple merger events in this cluster. In this paper, we analyze the substantial archival X-ray observations of A3667 from the Chandra X-ray Observatory and compare these with existing radio observations as well as state-of-the-art adaptive mesh refinement MHD cosmological simulations using Enzo. We have used two temperature map making techniques, weighted Voronoi tessellation and adaptive circular binning, to produce the high-resolution and largest field-of-view temperature maps of A3667. These high-fidelity temperature maps allow us to study the X-ray shocks in the cluster using a new two-dimensional shock-finding algorithm. We have also estimated the Mach numbers from the shocks inferred from previous ATCA radio observations. The combined shock statistics from the X-ray and radio data are in agreement with the shock statistics in a simulated MHD cluster. We have also studied the profiles of the thermodynamic properties across the cold front using ∼447 ks from the combined Chandra observations on A3667. Our results show that the stability of the cold front in A3667 can be attributed to the suppression of the thermal conduction across the cold front by a factor of ∼100-700 compared to the classical Spitzer value.

79 ASTRONOMY AND ASTROPHYSICS↗

Outlier classification using autoencoders: Application for fluctuation driven flows in fusion plasmas

Understanding the statistics of fluctuation driven flows in the boundary layer of magnetically confined plasmas is desired to accurately model the lifetime of the vacuum vessel components. Mirror Langmuir probes (MLPs) are a novel diagnostic that uniquely allow us to sample the plasma parameters on a time scale shorter than the characteristic time scale of their fluctuations. Sudden large-amplitude fluctuations in the plasma degrade the precision and accuracy of the plasma parameters reported by MLPs for cases in which the probe bias range is of insufficient amplitude. While some data samples can readily be classified as valid and invalid, we find that such a classification may be ambiguous for up to 40% of data sampled for the plasma parameters and bias voltages considered in this study. In this contribution to this work, we employ an autoencoder (AE) to learn a low-dimensional representation of valid data samples. By definition, the coordinates in this space are the features that mostly characterize valid data. Ambiguous data samples are classified in this space using standard classifiers for vectorial data. In this way, we avoid defining complicated threshold rules to identify outliers, which require strong assumptions and introduce biases in the analysis. By removing the outliers that are identified in the latent low-dimensional space of the AE, we find that the average conductive and convective radial heat fluxes are between approximately 5% and 15% lower as when removing outliers identified by threshold values. For contributions to the radial heat flux due to triple correlations, the difference is up to 40%

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Estimating Leaf Area Index in Row Crops Using Wheel-Based and Airborne Discrete Return Light Detection and Ranging Data

Leaf area index (LAI) is an important variable for characterizing plant canopy in crop models. It is traditionally defined as the total one-sided leaf area per unit ground area and is estimated by both direct and indirect methods. This paper explores the effectiveness of using light detection and ranging (LiDAR) data to estimate LAI for sorghum and maize with different treatments at multiple times during the growing season from both a wheeled vehicle and Unmanned Aerial Vehicles. Linear and nonlinear regression models are investigated for prediction utilizing statistical and plant structure-based features extracted from the LiDAR point cloud data with ground reference obtained from an in-field plant canopy analyzer (indirect method). Results based on the value of the coefficient of determination ( R 2 ) and root mean squared error for predictive models ranged from ∼0.4 in the early season to ∼0.6 for sorghum and ∼0.5 to 0.80 for maize from 40 Days after Sowing to harvest.

59 BASIC BIOLOGICAL SCIENCES↗

4D-STEM Coupled with Unsupervised Machine Learning to Reveal at Large-Scale the Microstructural Evolution in Li- and Mn-Rich Cathodes

Li- and Mn-rich (LMR) layered oxides are known to exhibit a thin surface reconstruction layer, which grows during electrochemical cycling in a manner that depends on exposed crystallographic facets, cycling conditions, and electrolyte chemistry. Direct characterization of this layer has traditionally relied on high-resolution electron microscopy, which is inherently limited to small fields of view. Here, we employ four-dimensional scanning transmission electron microscopy (4D-STEM) combined with unsupervised machine-learning clustering to quantitatively map phase distributions over large areas and track their evolution in LMR cathodes during electrochemical aging. Our results show that the surface reconstruction layer consists predominantly of a rocksalt phase, whose thickness varies across different facets following activation cycling and becomes substantially thicker and more uniform during calendar aging. In contrast, a spinel-like phase is observed within the particle bulk. Large-area phase mapping and correlative high-resolution imaging reveal that this spinel-like phase preferentially nucleates at bulk crystallographic defects, including boundaries between 60°-rotated layered domains and associated mixed-phase regions, rather than exclusively at the particle surface. Our findings establish a mechanistic distinction between surface-driven rocksalt formation and bulk-defect-mediated spinel nucleation while demonstrating the unique capability of 4D-STEM to provide statistically robust, mesoscale insight into complex phase-evolution processes in LMR cathodes.

4D-STEM↗