Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Labeled Datasets”

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 55 records · Page 3

Discovering Research Areas in Dataset Applications Through Knowledge Graphs and Large Language Models

Scientific datasets are increasingly cited in peer-reviewed journal publications, facilitating easy access to research utilizing those datasets. Datasets undergo a life cycle where older versions of datasets are replaced by newer versions often due to improvements in data resolution, algorithms, and other factors. Unlike peer reviewed documents registered with a single Digital Unique Identifier (DOI), datasets can be updated over time and the newer version of the datasets are registered with a new DOI which is not necessarily linked to the previous version of the dataset. It is challenging when publications citing a dataset need to be traced over the entire life cycle of that dataset. We provide an innovative approach to link the dataset versions and publications using a knowledge graph (KG). KG can help to trace the dataset cited in publications over the entire dataset life cycle and shed light into dataset usage in various applied research areas. We fine-tuned the pretrained NASA IMPACTINDUS Large Language Model (LLM) on a set of labeled publications abstracts. Our results showed that 87% of the publications were classified into one of twenty applied research areas, while the remaining 13% were classified into non-applied research areas. By linking datasets to applied research areas through the KG and employing Global Change Master Directory(GCMD), a well-established controlled vocabulary of scientific keywords describing Earth science datasets, we contribute to a transparent and advanced search and discovery mechanism for datasets across the Earth data ecosystem. The integrated KG and LLM approach is now incorporated and operational in dataset publication management at one of NASA’s Earth science data archival centers.

data provenance↗

A Cloud Detection Neural Network for Above-Aircraft Clouds Using Airborne Cameras

We introduce a method convolutional neural networks to detect the presence of clouds in airborne camera images. We quantify the performance of this Cloud Detection Neural Network (CDNN) using human-labeled validation data where we report a 96% accuracy in detecting clouds in testing datasets for both Zenith- viewing and Forward-viewing models. We assess our performance by comparing the flight-averaged cloud fraction of zenith and forward CDNN retrievals, with that of the prototype hyperspectral total-diffuse Sunshine Pyranometer (SPN-S) instrument’s cloud optical depth data. Comparison of the CDNN with the SPN-S on time specific intervals resulted in 93% accuracy for the zenith- viewing CDNN and 84% for the forward- viewing CDNN. The comparison of the CDNNs with the SPN-S on flight-averaged cloud fraction resulted in an agreement of 0.15 for the Forward CDNN and 0.07 for the Zenith CDNN. We then quantify the ability of the CDNN to identify the presence of clouds above the aircraft using a forward- looking camera mounted inside the aircraft cockpit compared to the use of an All- Sky upward-looking camera that is mounted outside the fuselage on top of the aircraft. We present results from the CDNN based on airborne imagery from the NASA Aerosol Cloud Meteorology Interactions Over the Western Atlantic Experiment (ACTIVATE) and the Clouds, Aerosol and Monsoon Processes- Philippines Experiment (CAMP2Ex). For CAMP2Ex 53% of flight dates had above- aircraft cloud fraction above 50%, while for ACTIVATE 52% and 54% of flight dates observed above-aircraft cloud fraction above 50% for 2020 and 2021, respectively.

Joseph D Nied↗

LLMs and GenAI Tools to Depict Contributions of Human Systems to Spaceflight Tasks Execution

Recent advancements in Artificial Intelligence and Machine Learning (AI/ML) technologies, particularly Large Language Models (LLMs) capable of sophisticated syntax analysis, offer substantial potential in automating complex processes, thereby saving time and human resources. This study explores the development of an LLM-driven model designed to analyze and categorize a diverse set of Mars mission tasks into 18 predefined Human System Task Categories (HSTCs) based on their textual descriptions. As part of developing the Crew Health and Performance – Probabilistic Risk Assessment (CHP-PRA projects Performance Risk Model (PRisM) proof-of-concept, we established a framework to project performance scores from small-scale tests onto a preliminary list of Mars tasks. The foundation of our model was a comprehensive spreadsheet populated by NASA experts and clinicians, which detailed each Mars task alongside binary indicators of HSTC involvement. This dataset enabled the initial application of supervised ML, training and testing on existing HSTC labels. The HSTCs were originally defined from a medical system perspective, focusing on task impairments due to deteriorated human health. To expand our model's scope to include categories impacting performance, we face the challenge of generating binary labels (0 or 1) for new categories without pre-existing data. We address this by employing Generative AI (GenAI) software to determine whether a given task involved a new category by asking, "Does task A involve using category B?" We validate our approach by comparing the GenAI's binary classifications with the expert-provided labels for existing HSTCs. Notably, we utilize Ollama [4], a locally hosted GenAI tool that does not require cloud access, thus safeguarding NASA's proprietary data from unauthorized exposure. This study demonstrates the feasibility of leveraging cutting-edge AI tools to advance research, paving the way for automation and rapid decision-making in space exploration.

Mona Matar↗

Curating AI-Ready Datasets for Equity and Environmental Justice: A Data-Centric AI Case Study

An equitable and environmentally just community is essentialin order to avoid disproportionate burden borne by vulnerablecommunities. This need becomes pressing in the aftermathof an extreme event such as disaster or hazard when it is diffi-cult for the governing bodies to implement resource allocationas per the need. Artificial Intelligence (AI) algorithms canhelp surface Equity and Environmental Justice (EEJ) issueswhen trained on EEJ datasets. However, curating AI-readyEEJ training datasets is challenging due to differences in fac-tors such as heterogeneity, resolution, modality, and level ofexpertise in labeling. Additionally, EEJ issues involve sensi-tive information where uncertainties and errors could degradethe performance of AI algorithms. For eg. Error in seasonalcrop yield information can highly affect the prediction of an-nual crop yield. To address these challenges, Data-centricAI (DCAI) methods are employed, which enhance AI algo-rithm performance even with limited training samples. DCAIprioritizes data quality, thereby reducing the adverse effectsof uncertainties and errors during the model training process.This research proposes a novel dataset and benchmark for an-alyzing the effect of the Maui Wildfire of 2023 for Equityand Environmental Justice (EEJ) issues. The proposed datasetaligns with the concepts of DCAI such as annotation quality,data preprocessing, privacy, feature engineering, governanceand provenance. We firmly believe that the proposed datasetwould lay a foundation to implement robust and reliable mod-ern AI algorithms for addressing EEJ issues.

Paridhi Parajuli↗

Deep Domain Adaptation based Cloud Type Detection using Active and Passive Satellite Data

Domain adaptation techniques have been developed to handle data from multiple sources or domains. Most existing domain adaptation models assume that source and target domains are homogeneous, i.e., they have the same feature space. Nevertheless, many real world applications often deal with data from heterogeneous domains that come from completely different feature spaces. In our remote sensing application, data in source domain (from an active spaceborne Lidar sensor CALIOP onboard CALIPSO satellite) contain 25 attributes, while data in target domain (from a passive spectroradiometer sensor VIIRS onboard Suomi-NPP satellite) contain 20 different attributes. CALIOP has better representation capability and sensitivity to aerosol types and cloud phase, while VIIRS has wide swaths and better spatial coverage but has inherent weakness in differentiating atmospheric objects on different vertical levels. To address this mismatch of features across the domains/sensors, we propose a novel end-to-end deep domain adaptation with domain mapping and correlation alignment (DAMA) to align the heterogeneous source and target domains in active and passive satellite remote sensing data. It can learn domain invariant representation from source and target domains by transferring knowledge across these domains, and achieve additional performance improvement by incorporating weak label information into the model (DAMA-WL). Our experiments on a collocated CALIOP and VIIRS dataset show that DAMA and DAMA-WL can achieve higher classification accuracy in predicting cloud types.

domain adaptation↗

Enhancing Dataset Discovery and Usage Tracking in Earth Sciences: Integrating Knowledge Graphs and Large Language Models

NASA's Data Active Archive Centers (DAACs) have played a crucial role in supporting a wide range of applied research in Earth and Environmental sciences. To date, over 20,000 publications have been collected, citing more than 3,000 NASA Earth science datasets. We present an innovative approach that links datasets and collected publications through a knowledge graph (KG). This KG enables the tracking of dataset citations throughout the dataset's lifecycle, revealing patterns of dataset usage across various applied research areas. We fine-tuned the pre-trained NASA IMPACT INDUS-Base Retriever Large Language Model (LLM) using a set of labeled publication abstracts. Our results indicate that 87% of the publications were classified into one of twenty applied research areas, while the remaining 13% were categorized into non-applied research areas. The classified publications linked to datasets are used to discover datasets by users interested in specific applied research and by dataset providers to determine dataset usage for applications.

open-source↗

A 3D Active Learning Application for NeMO-Net, the NASA Neural Multi-Modal Observation and Training Network for Global Coral Reef Assessment

NeMO-Net, the NASA neural multi-modal observation and training network for global coral reef assessment, is an open-source deep convolutional neural network and interactive active learning training software aiming to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology as well as mapping of spatial distribution. We present an interactive video game prototype for tablet and mobile devices where users interactively label morphology classifications over mm-scale 3D coral reef imagery captured using fluid lensing to create a dataset that will be used to train NeMO-Nets convolutional neural network. The application currently allows for users to classify preselected regions of coral in the Pacific and will be expanded to include additional regions captured using our NASA FluidCam instrument, presently the highest-resolution remote sensing benthic imaging technology capable of removing ocean wave distortion, as well as lower-resolution airborne remote sensing data from the ongoing NASA CORAL campaign. Active learning applications present a novel methodology for efficiently training large-scale Neural Networks wherein variances in identification can be rapidly mitigated against control data. NeMO-Net periodically checks users input against pre-classified coral imagery to gauge their accuracy and utilize in-game mechanics to provide classification training. Users actively communicate with a server and are requested to classify areas of coral for which other users had conflicting classifications and contribute their input to a larger database for ranking. In partnering with Mission Blue and IUCN, NeMO-Net leverages an international consortium of subject matter experts to classify areas of confusion identified by NeMO-Net and generate additional labels crucial for identifying decision boundary locations in coral reef assessment.

NeMO=NET↗

A Machine Learning-Based Cloud Detection and Thermodynamic Phase Classification Algorithm using Passive Spectral Observations

We trained two Random Forest (RF) machine-learning models for cloud mask and cloud thermodynamic phase detection using spectral observations from VIIRS on Suomi NPP (SNPP). Observations from CALIOP were carefully selected to provide reference labels. The two RF models were trained for all-day and daytime-only conditions using a 4-year collocated VIIRS/CALIOP dataset from 2013 to 2016. Due to the orbit difference, the collocated CALIOP and SNPP VIIRS training samples cover a broad viewing zenith angle range, which is a great benefit to overall model performance. The all-day model uses 3 VIIRS infrared (IR) bands (8.6,11, and 12 μm) and the daytime model uses 5 Near-IR (NIR) and Shortwave-IR (SWIR) bands (0.86, 1.24, 1.38, 1.64 and 2.25 μm) together with the 3 IR bands to detect clear, liquid water, and ice cloud pixels. Up to 7 surface types, namely, ocean/water, forest, cropland, grassland, snow/ice, barren/desert, and shrubland, were considered separately to enhance performance for both models. Detection of cloudy pixels and thermodynamic phase with the two RF models were compared against collocated CALIOP products from 2017. It is shown that, with a conservative screening process that excludes the most challenging cloudy pixels for passive remote sensing, the two RF models have high accuracy rates in comparison with the CALIOP reference for both cloud detection and thermodynamic phase. Other existing SNPP VIIRS and Aqua MODIS cloud mask and phase products are also evaluated, with results showing that the two RF models and the MODIS MYD06 optical property phase product are the top 3 algorithms with respect to lidar observations during the daytime. During the nighttime, the RF all-day model works best for both cloud detection and phase, in particular for pixels over snow/ice surfaces. The present RF models can be extended to other similar passive instruments if training samples can be collected from CALIOP or other lidars. However, the quality of reference labels and potential sampling issues that may impact model performance would need further attention.

cloud detection↗

ImageLabler: Labeling and Managing Image Data for Machine Learning in the Earth Sciences

While machine learning techniques for image classification have been around for a long time, storing and managing the vast number of images required as training data is still a problem for scientists. This is especially true for the field of Earth science, where only recently have experts begun using machine learning techniques for image-based phenomena classification. Image Labeler, a fast and scalable cloud-based tagging platform for Earth science images, seeks to improve upon existing methods of managing images and associated metadata, such as maintaining categorized folders of images on a local machine, a process that can be cumbersome and difficult to scale. The platform facilitates rapid development of image-based Earth science phenomena training datasets by allowing scientists to upload their existing imagery as well as extract new samples from open satellite imagery services made available through NASA’s Global Imagery Browse Service (GIBS). Image Labeler also supports GeoTIFF data, with capabilities such as displaying GeoTIFFs on an interactive map, drawing shapefiles over them, and tagging them with additional metadata. This allows scientists to perform spatiotemporal subsetting with geographic information and develop training data more quickly. Built using modern web technologies, Image Labeler includes additional capabilities such as team collaboration for large-scale image tagging projects. Users can download their data in a machine-learning-ready format, allowing scientists to spend time on experimentation rather than on the collection of training data. In this presentation, we demonstrate how Image Labeler seeks to become a one-stop image data management solution for machine learning applications in Earth science.

Ashish Acharya↗

Improving Trust in Deep Neural Networks with Nearest Neighbors

Deep neural networks are used increasingly for perception and decision-making in UAVs. For example, they can be used to recognize objects from images and decide what actions the vehicle should take. While deep neural networks can perform very well at complex tasks, their decisions may be unintuitive to a human operator. When a human disagrees with a neural network prediction, due to the black box nature of deep neural networks, it can be unclear whether the system knows something the human does not or whether the system is malfunctioning. This uncertainty is problematic when it comes to ensuring safety. As a result, it is important to develop technologies for explaining neural network decisions for trust and safety. This paper explores a modification to the deep neural network classification layer to produce both a predicted label and an explanation to support its prediction. Specifically, at test time, we replace the final output layer of the neural network classifier by a k-nearest neighbor classifier. The nearest neighbor classifier produces 1) a predicted label through voting and 2) the nearest neighbors involved in the prediction, which represent the most similar examples from the training dataset. Because prediction and explanation are derived from the same underlying process, this approach guarantees that the explanations are always relevant to the predictions. We demonstrate the approach on a convolutional neural network for a UAV image classification task. We perform experiments using a forest trail image dataset and show empirically that the hybrid classifier can produce intuitive explanations without loss of predictive performance compared to the original neural network. We also show how the approach can be used to help identify potential issues in the network and training process.

Lee, Ritchie↗

Digitizing Named Entities Found Within Letters of Agreement

Letters of Agreement (LOAs) are text-based air traffic control documents that contain procedures and actions agreed upon by the different parties, typically two or more FAA facilities, that are subject to an agreement. The documents contain among other things generic constraints, which are explicit and implicit combinations of procedures that limit a flight’s trajectory and affects pilot actions. For example, a controller may be required, to assign a specific altitude to an aircraft crossing the boundary between two airspaces. Although LOA generic constraints directly impact the trajectory of an aircraft, they are not currently available in a digital form that can be used for (or directly ingested into automated) flight planning. Instead, the constraints are manually input into an onboard or ground based system. LOA documents are primarily stored at a controlling facility and the generic constraints are implemented by experienced air traffic controllers and pilots primarily using voice instructions. This increases the workload of the controllers, likelihood of error (e.g., due to noisy communication) and makes it impractical for implementation with unmanned aircraft. Therefore, steps must be taken to make existing constraints machine interpretable to enable e.g., automated handoffs which in turn would reduce controller workload. With recent advances in natural language processing, especially the rise in digitization of text documents (e.g., medical documents) and automated extraction of information therein, it is now possible to extract flight specific constraints from LOAs. The goal of this work is to digitize named entities through a combination of natural language processing tasks: named entity disambiguation, toponym resolution, and numeric parsing to extract general constraint components contained within LOAs, herein referred to as Entity Enhancement (EE). Starting with a small list of named entities (e.g., ARTCC, Tower, Altitude and Speed), EE can extract the named entities while simultaneously converting the string-based output into a digital format using an ensemble of processes like rule-based gazetteers and syntactic-lexical patterns. The digital format contains a diverse set of information based on the entity label in question, ranging from standardized facility names to units of measure (e.g., feet) and other numeric information. Upon validating our approach using a truth dataset, we show an overall F1-Score of 0.71 for the extraction process. Looking beyond entity enhancement, we are also working towards the goal of completely digitizing the general constraints by performing EE and fitting them into a standardized exchange model (XM) such as the Aeronautical Information Exchange Model (AIXM). This will allow for easy distribution and dissemination of LOA constraints to air users, better searchability within documents, and enable ingestion into automated flight planning. Finally, we show a preliminary version of the proposed XM architecture and demonstrate how the model can be populated from the EE output.

Stephen S. B. Clarke↗

Water Across Synthetic Aperture Radar Data (WASARD): SAR Water Body Classification for the Open Data Cube

The detection of inland water bodies from Synthetic Aperture Radar (SAR) data provides a great advantage over water detection with optical data, since SAR imaging is not impeded by cloud cover. Traditional methods of detecting water from SAR data involves using thresholding methods that can be labor intensive and imprecise. This paper describes Water Across Synthetic Aperture Radar Data (WASARD): a method of water detection from SAR data which automates and simplifies the thresholding process using machine learning on training data created from Geoscience Australia’s WOFS algorithm. Of the machine learning models tested, the Linear Support Vector Machine was determined to be optimal, with the option of training using solely the VH polarization or a combination of the VH and VV polarizations. WASARD was able to identify water in the target area with a correlation of 97% with WOFS. Sentinel-1, Open Data Cube, Earth Observations, Machine Learning, Water Detection 1. INTRODUCTION Water classification is an important function of Earth imaging satellites, as accurate remote classification of land and water can assist in land use analysis, flood prediction, climate change research, as well as a variety of agricultural applications [2]. The ability to identify bodies of water remotely via satellite is immensely cheaper than contracting surveys of the areas in question, meaning that an application that can accurately use satellite data towards this function can make valuable information available to nations which would not be able to afford it otherwise. Highly reliable applications for the remote detection of water currently exist for use with optical satellite data such as that provided by LANDSAT. One such application, Geoscience Australia’s Water Observations from Space (WOFS) has already been ported for use with the Open Data Cube [6]. However, water detection using optical data from Landsat is constrained by its relatively long revisit cycle of 16 days [5], and water detection using any optical data is constrained in that it lacks the ability to make accurate classifications through cloud cover [2]. The alternative solution which solves these problems is water detection using SAR data, which images the Earth using cloud-penetrating microwaves. Because of its advantages over optical data, much research has been done into water detection using SAR data. Traditionally, this has been done using the thresholding method, which involves picking a polarization band and labeling all pixels for which this band’s value is below a certain threshold as containing water. The thresholding method works since water tends to return a much lower backscatter value to the satellite than land [1]. However, this method can be flawed since estimating the proper threshold is often imprecise, complicated, and labor intensive for the end user. Thresholding also tends to use data from only one SAR polarization, when a combination of polarizations can provide insight into whether water is present. [2] In order to alleviate these problems, this paper presents an application for the Open Data Cube to detect water from SAR data using support vector machine (SVM) classification. 2. PLATFORM WASARD is an application for the Open Data Cube, a mechanism which provides a simple yet efficient means of ingesting, storing, and retrieving remote sensing data. Data can be ingested and made analysis ready according to whatever specifications the researcher chooses, and easily resampled to artificially alter a scene’s resolution. Currently WASARD supports water detection on scenes from ESA’s Sentinel-1 and JAXA’s ALOS. When testing WASARD, Sentinel-1 was most commonly used due to its relatively high spatial resolution and its rapid 6 day revisit cycle [5]. With minor alterations to the application's code, however, it could support data from other satellites. 3. METHODOLOGY Using supervised classification, WASARD compares SAR data to a dataset pre-classified by WOFS in order to train an SVM classifier. This classifier is then used to detect water in other SAR scenes outside the training set. Accuracy was measured according to the following metrics:  Precision: a measure of what percentage of the points WASARD labels as water are truly water  Recall: a measure of what percentage of the total water cover WASARD was able to identify.  F1 Score: a harmonic average of the precision and recall scores Both precision and recall are calculated at the end of the training phase, when the trained classifier is compared to a testing dataset. Because the WOFS algorithm’s classifications are used as the truth values when training a WASARD classifier, when precision and recall are mentioned in this paper, they are always with respect to the values produced by WOFS on a similar scene of Landsat data, which themselves have a classification accuracy of 97% [6]. Visual representations of water identified by WASARD in this paper were produced using the function wasard_plot(), which is included in WASARD. 3.1 Algorithm Selection The machine learning model used by WASARD is the Linear Support Vector Machine (SVM). This model uses a supervised learning algorithm to develop a classifier, meaning it creates a vector which can be multiplied by the vector formed by the relevant data bands to determine whether a pixel in a SAR scene contains water. This classifier is trained by comparing data points from selected bands in a SAR scene to their respective labels, which in this case are “water” or “not water” as given by the WOFS algorithm. The SVM was selected over the Random Forest model, which outperformed the SVM in training speed, but had a greater classification time and lower accuracy, and the Multilayer Perceptron Artificial Neural Network, which had a slightly higher average accuracy than the SVM, but much greater training and classification times. Figure 1: Visual representation of the SVM Classifier. Each white point represents a pixel in a SAR scene. In Figure 1, the diagonal line separating pixels determined to be water from those determined not to be water represents the actual classification vector produced by the SVM. It is worth noting that once the model has been trained, classification of pixels is done in a similar manner as in the thresholding method. This is especially true if only one band was used to train the model. 3.1 Feature Selection Sentinel-1 collects data from two bands: the Vertical/Vertical polarization (VV) and the Vertical/Horizontal polarization (VH). When 100 SVM classifiers were created for each polarization individually, and for the combination of the two, the following results were achieved: Figure 2: Accuracy of classifiers trained using different polarization bands. Precision and Recall were measured with respect to the values produced by WOFS. Figure 2 demonstrates that using both the VV and VH bands trades slightly lower recall for significantly greater precision when compared with the VH band alone, and that using the VV band alone is inferior in both metrics. WASARD therefore defaults to using both the VV and VH bands, and includes the option to use solely the VH band. The VV polarization’s lower precision compared to the VH polarization is in contrast to results from previous research and may merit further analysis [4]. 3.2 Training a Classifier The steps in training a classifier with WASARD are 1. Selecting two scenes (one SAR, one optical) with the same spatial extents, and acquired close to each other in time, with a preference that the scenes are taken on the same day. 2. Using the WOFS algorithm to produce an array of the detected water in the scene of optical data, to be used as the labels during supervised learning 3. Data points from the selected bands from the SAR acquisition are bundled together into an array with the corresponding labels gathered from WOFS. A random sample with an equal number of points labeled “Water” and “Not Water” is selected to be partitioned into a training and a testing dataset 4. Using Scikit-Learn’s LinearSVC object, the training dataset is used to produce a classifier, which is then tested against the testing dataset to determine its precision and recall The result is a wasard_classifier object, which has the following attributes: 1. f1, recall, and precision: 3 metrics used to determine the classifier’s accuracy 2. Coefficient: Vector which the SVM uses to make its predictions. The classifier detects water when the dot product of the coefficient and the vector formed by the SAR bands is positive 3. Save(): allows a user to save a classifier to the disk in order to use it without retraining 4. wasard_classify(): Classifies an entire xarray of SAR data using the SVM classifier All of the above steps are performed automatically when the user creates a wasard_classifier object. 3.3 Classifying a Dataset Once the classifier has been created, it can be used to detect water in an xarray of SAR data using wasard_classify(). By taking the dot product of the classifier’s coefficients and the vector formed by the selected bands of SAR data, an array of predictions is constructed. A classifier can effectively be used on the same spatial extents as the ones where it was trained, or on any area with a similar landscape. While

Kreiser, Zachary↗

Identifying Emerging Safety Threats Through Topic Modeling in the Aviation Safety Reporting System: A Covid-19 Study

The NASA Aviation Safety Reporting System (ASRS) is a voluntary, confidential aviation safety reporting system. The ASRS receives reports from pilots, air traffic controllers, flight attendants, and others involved in aviation operations. The reports are de-identified and coded by ASRS expert safety analysts, and a short descriptive synopsis is written to describe the safety issue. The de-identified reports are then disseminated to the aviation community in many ways, including via an online database, Safety Alert Bulletins, For Your Information Notices, and the CALLBACK newsletter. In this work, we consider whether we can improve the grouping, linking, and understanding of safety concerns through topic modeling. Specifically, we use topic modeling as a building block to identify emerging safety threats over time. This unsupervised approach, we argue, offers the flexibility to identify new emerging themes in this large dataset by constructing different timelines based on the content similarity of ASRS report narratives. This method's unsupervised nature improves upon related research, which is limited to pre-defined labels and therefore can not fully capture emerging safety threats. We apply our method to all ASRS reports in 2020 to assess if the generated timelines can highlight COVID-19 as it is emerging as a safety threat in incoming ASRS reports. We perform both a quantitative and qualitative evaluation of the automatically constructed timelines. The qualitative evaluation is performed by describing the evolution of top terms in the timelines, generated by our method, which we found explicitly convey the themes of COVID-19. Separately, we use a set of 1,213 COVID-19 reports from 2020 that were manually identified by ASRS analysts to quantitatively evaluate the COVID-19 reports distribution across the timelines. Our results have shown that COVID-19 emergence can be identified using the top terms that were generated by topic modeling. The top terms in topic modeling therefore can serve as a summary alternative to manually inspecting reports. Moreover, leveraging the manually identified COVID-19 reports, we found the manually identified timelines accounted for over 70% of the COVID-19 reports curated by the ASRS analysts, which demonstrates the potential of this approach for facilitating the understanding of safety concerns as they emerge and evolve. This method shows great potential to understand aerospace safety threats and other narrative- driven incident report databases.

ASRS↗

Identifying Emerging Safety Threats Through Topic Modeling in the Aviation Safety Reporting System: A COVID-19 Study

The NASA Aviation Safety Reporting System (ASRS) is a voluntary, confidential aviation safety reporting system. The ASRS receives reports from pilots, air traffic controllers, flight attendants, and others involved in aviation operations. The reports are de-identified and coded by ASRS expert safety analysts, and a short descriptive synopsis is written to describe the safety issue. The de-identified reports are then disseminated to the aviation community in many ways, including via an online database, Safety Alert Bulletins, For Your Information Notices, and the CALLBACK newsletter. In this work, we consider whether we can improve the grouping, linking, and understanding of safety concerns through topic modeling. Specifically, we use topic modeling as a building block to identify emerging safety threats over time. This unsupervised approach, we argue, offers the flexibility to identify new emerging themes in this large dataset by constructing different timelines based on the content similarity of ASRS report narratives. This method's unsupervised nature improves upon related research, which is limited to pre-defined labels and therefore can not fully capture emerging safety threats. We apply our method to all ASRS reports in 2020 to assess if the generated timelines can highlight COVID-19 as it is emerging as a safety threat in incoming ASRS reports. We perform both a quantitative and qualitative evaluation of the automatically constructed timelines. The qualitative evaluation is performed by describing the evolution of top terms in the timelines, generated by our method, which we found explicitly convey the themes of COVID-19. Separately, we use a set of 1,213 COVID-19 reports from 2020 that were manually identified by ASRS analysts to quantitatively evaluate the COVID-19 reports distribution across the timelines. Our results have shown that COVID-19 emergence can be identified using the top terms that were generated by topic modeling. The top terms in topic modeling therefore can serve as a summary alternative to manually inspecting reports. Moreover, leveraging the manually identified COVID-19 reports, we found the manually identified timelines accounted for over 70% of the COVID-19 reports curated by the ASRS analysts, which demonstrates the potential of this approach for facilitating the understanding of safety concerns as they emerge and evolve. This method shows great potential to understand aerospace safety threats and other narrative- driven incident report databases.

ASRS↗

Towards Physics Guided Optical Flow for Tracking Atmospheric Motion

Atmospheric 3D winds in the horizontal and vertical directions are critical for improving short-range and long-range forecasting. Such advancement in forecasting directly applies to research in a number of areas including convective processes, wildfire plumes and tornado prediction. Atmospheric Motion Vectors (AMVs) provide a passively sensed approach to quantifying horizontal motion and cloud heights, which are typically sourced from geostationary sensors due to the availability of high frequency observations. Recent work has shown that estimating AMVs by tracking individual pixels with dense optical flow is a promising new direction. In this work, we use a state-of-the-art convolutional neural network for optical flow (FlowNetS) in a physics-guided deep learning framework for predicting AMVs in the horizontal direction. The approach is semi-supervised and uses physically informed wind vectors from high-resolution numerical simulations (DYAMOND) for supervised learning followed by fine-tuning though warping and reconstruction of full-disk geostationary images (GOES-16). In the vertical direction, we use labels from the CALIPSO low-earth orbit satellite to predict cloud height from 16-band geostationary images with a neural network. We present results for both tasks on held-out time periods and secondary datasets.

geostationary↗

Towards Physics Guided Optical Flow for Tracking Atmospheric Motion

Observations of atmospheric 3D winds are critical for improving short-range and long-range forecasting. Such advancement in forecasting directly applies to research in a number of areas including convective processes, wildfire plumes and tornado prediction. Atmospheric Motion Vectors (AMVs) provide a passively sensed approach to quantifying horizontal motion and cloud heights, which are typically sourced from geostationary sensors due to the availability of high frequency observations. Recent work has shown that estimating AMVs by tracking individual pixels with dense optical flow is a promising new direction. In this work, we compare state-of-the-art convolutional neural networks for optical flow in a physics-guided deep learning framework for predicting AMVs. The approach is semi-supervised and uses physically informed wind vectors from high-resolution numerical simulations for supervised learning followed by fine-tuning though warping and reconstruction of full-disk geostationary images (GOES-16). In the vertical direction, we use labels from the CALIPSO low-earth orbit satellite to predict cloud height from 16-band geostationary images with a neural network. We present results for both tasks on held-out time periods and secondary datasets.

Geostationary↗

A Global Land Cover Training Dataset From 1984 to 2020

State-of-the-art cloud computing platforms such as Google Earth Engine (GEE) enable regional-to-global land cover and land cover change mapping with machine learning algorithms. However, collection of high-quality training data, which is necessary for accurate land cover mapping, remains costly and labor-intensive. To address this need, we created a global database of nearly 2 million training units spanning the period from 1984 to 2020 for seven primary and nine secondary land cover classes. Our training data collection approach leveraged GEE and machine learning algorithms to ensure data quality and biogeographic representation. We sampled the spectral-temporal feature space from Landsat imagery to efficiently allocate training data across global ecoregions and incorporated publicly available and collaborator-provided datasets to our database. To reflect the underlying regional class distribution and post-disturbance landscapes, we strategically augmented the database. We used a machine learning-based cross-validation procedure to remove potentially mis-labeled training units. Our training database is relevant for a wide array of studies such as land cover change, agriculture, forestry, hydrology, urban development, among many others.

Radost Stanimirova↗

Empirical Analysis and Automated Classification of Security Bug Reports

With the ever expanding amount of sensitive data being placed into computer systems, the need for effective cybersecurity is of utmost importance. However, there is a shortage of detailed empirical studies of security vulnerabilities from which cybersecurity metrics and best practices could be determined. This thesis has two main research goals: (1) to explore the distribution and characteristics of security vulnerabilities based on the information provided in bug tracking systems and (2) to develop data analytics approaches for automatic classification of bug reports as security or non-security related. This work is based on using three NASA datasets as case studies. The empirical analysis showed that the majority of software vulnerabilities belong only to a small number of types. Addressing these types of vulnerabilities will consequently lead to cost efficient improvement of software security. Since this analysis requires labeling of each bug report in the bug tracking system, we explored using machine learning to automate the classification of each bug report as a security or non-security related (two-class classification), as well as each security related bug report as specific security type (multiclass classification). In addition to using supervised machine learning algorithms, a novel unsupervised machine learning approach is proposed. An ac- curacy of 92%, recall of 96%, precision of 92%, probability of false alarm of 4%, F-Score of 81% and G-Score of 90% were the best results achieved during two-class classification. Furthermore, an accuracy of 80%, recall of 80%, precision of 94%, and F-score of 85% were the best results achieved during multiclass classification.

Cybersecurity↗