Results for “training data” — Page 62

NASA NTRS2011

Accelerated Aging Experiments for Prognostics of Damage Growth in Composite Materials

Composite structures are gaining importance for use in the aerospace industry. Compared to metallic structures their behavior is less well understood. This lack of understanding may pose constraints on their use. One possible way to deal with some of the risks associated with potential failure is to perform in-situ monitoring to detect precursors of failures. Prognostic algorithms can be used to predict impending failures. They require large amounts of training data to build and tune damage model for making useful predictions. One of the key aspects is to get confirmatory feedback from data as damage progresses. These kinds of data are rarely available from actual systems. The next possible resource to collect such data is an accelerated aging platform. To that end this paper describes a fatigue cycling experiment with the goal to stress carbon-carbon composite coupons with various layups. Piezoelectric disc sensors were used to periodically interrogate the system. Analysis showed distinct differences in the signatures of growing failures between data collected at conditions. Periodic X-radiographs were taken to assess the damage ground truth. Results after signal processing showed clear trends of damage growth that were correlated to damage assessed from the X-ray images.

composites↗

NASA NTRS2014

Artificial Intelligence Based Control Power Optimization on Tailless Aircraft

Traditional methods of control allocation optimization have shown difficulties in exploiting the full potential of controlling large arrays of control devices on innovative air vehicles. Artificial neutral networks are inspired by biological nervous systems and neurocomputing has successfully been applied to a variety of complex optimization problems. This project investigates the potential of applying neurocomputing to the control allocation optimization problem of Hybrid Wing Body (HWB) aircraft concepts to minimize control power, hinge moments, and actuator forces, while keeping system weights within acceptable limits. The main objective of this project is to develop a proof-of-concept process suitable to demonstrate the potential of using neurocomputing for optimizing actuation power for aircraft featuring multiple independently actuated control surfaces. A Nastran aeroservoelastic finite element model is used to generate a learning database of hinge moment and actuation power characteristics for an array of flight conditions and control surface deflections. An artificial neural network incorporating a genetic algorithm then uses this training data to perform control allocation optimization for the investigated aircraft configuration. The phase I project showed that optimization results for the sum of required hinge moments are improved by more than 12% over the best Nastran solution by using the neural network optimization process.

Gern, Frank↗

NASA NTRS2016

Compiling a Comprehensive EVA Training Dataset for NASA Astronauts

Training for a spacewalk or extravehicular activity (EVA) is considered hazardous duty for NASA astronauts. This activity places astronauts at risk for decompression sickness as well as various musculoskeletal disorders from working in the spacesuit. As a result, the operational and research communities over the years have requested access to EVA training data to supplement their studies.

Laughlin, M. S.↗

NASA NTRS2019

Application of High-Dimensional Fuzzy K-Means Cluster Analysis to CALIOP/CALIPSO Version 4.1 Cloud-Aerosol Discrimination

This study applies fuzzy k-means (FKM) cluster analyses to a subset of the parameters reported in the CALIPSO lidar level 2 data products in order to classify the layers detected as either clouds or aerosols. The results obtained are used to assess the reliability of the cloud–aerosol discrimination (CAD) scores reported in the version 4.1 release of the CALIPSO data products. FKM is an unsupervised learning algorithm, whereas the CALIPSO operational CAD algorithm (COCA) takes a highly supervised approach. Despite these substantial computational and architectural differences, our statistical analyses show that the FKM classifications agree with the COCA classifications for more than 94 % of the cases in the troposphere. This high degree of similarity is achieved because the lidar-measured signatures of the majority of the clouds and the aerosols are naturally distinct, and hence objective methods can independently and effectively separate the two classes in most cases. Classification differences most often occur in complex scenes (e.g., evaporating water cloud filaments embedded in dense aerosol) or when observing diffuse features that occur only intermittently (e.g., volcanic ash in the tropical tropopause layer). The two methods examined in this study establish overall classification correctness boundaries due to their differing algorithm uncertainties. In addition to comparing the outputs from the two algorithms, analysis of sampling, data training, performance measurements, fuzzy linear discriminants, defuzzification, error propagation, and key parameters in feature type discrimination with the FKM method are further discussed in order to better understand the utility and limits of the application of clustering algorithms to space lidar measurements. In general, we find that both FKM and COCA classification uncertainties are only minimally affected by noise in the CALIPSO measurements, though both algorithms can be challenged by especially complex scenes containing mixtures of discrete layer types. Our analysis results show that attenuated backscatter and color ratio are the driving factors that separate water clouds from aerosols; backscatter intensity, depolarization, and mid-layer altitude are most useful in discriminating between aerosols and ice clouds; and the joint distribution of backscatter intensity and depolarization ratio is critically important for distinguishing ice clouds from water clouds.

Zeng, Shan↗

NASA NTRS2019

Utilizing Convolutional Neural Networks for Global Seagrass Habitat Mapping

Convolutional neural networks (CNNs) are becoming an increasingly prevalent machine learning algorithm due to their high accuracy and lack of reliance on heuristic processes. One of the major drawbacks of convolutional neural networks is their reliance on large amounts of training data in order to generate sensible results. This talk will cover how our team has utilized the strengths and overcome the weaknesses of convolutional neural networks as they apply to seagrass habitat mapping. We will share our technical CNN results over time, detail the requirements and challenges that our team overcame and explore how other teams can better incorporate a stronger seagrass component into their machine learning projects.

Convolutional↗

NASA NTRSDate not supplied

TPSAS-NF1676L-26990-DND

The warm phase ENSO event (El Nino) of 2015/2016 produced significant perturbations to the Tropical Pacific, from sea surface to tropical tropopause and above. Although there have been 3 El Ninos during the modern "A-Train" era, the most recent El Nino started earlier in the year, was located farther west and was stronger than those of 2006/7 and 2009/10. Using A-Train observations, we show that during the 2015/16 El Nino the center of convective activity moved from the tropical western Pacific (TWP) to the central Pacific, just to the east of the dateline. This caused a record amount of cloud ice to occur near or above the environmental cold-point tropopause. Coincidently there was a large positive anomaly in tropical lower stratospheric water vapor, warming of the tropopause over the TWP, and changes to the stratospheric QBO. Using ten+ years of A-Train data and the MERRA-2 reanalysis fields, we explore the observed relationships between the 3D patterns of cloud ice, water vapor, ozone, aerosols and other trace gases in the tropical Pacific, and correlations with the ENSO and QBO cycles.

Melody A. Avery↗

NASA NTRSDate not supplied

Human-system Interaction Issues and Proposed Solutions To Promote Successful Maturation of the UTM System

Over five years, NASA, together with partnering organizations, has been developing and successfully demonstrating the maturing capabilities of the Unmanned Aircraft Systems (UAS) Traffic Management (UTM) system and its ability to support communication and coordination among small UAS operations through a series of flight tests. During these flight tests, human-system interaction (HSI) elements were also explored in order to identify the barriers to implementation as human operators transitionally fulfill roles that will be ultimately tasked to future automation. Throughout the tests, similar issues were regularly documented and are expected to persist if not formally addressed by consistent procedures or regulation. Documented here, along with suggested mitigations, are the most frequently noted HSI items, which include operator training, data standardization, and information quantity.

UTM↗

NASA NTRSDate not supplied

Human-system interaction issues and proposed solutions to promote successful maturation of the UTM system

Over five years, NASA, together with partnering organizations, has been developing and successfully demonstrating the maturing capabilities of the Unmanned Aircraft Systems (UAS) Traffic Management (UTM) system and its ability to support communication and coordination among small UAS operations through a series of flight tests. During these flight tests, human-system interaction (HSI) elements were also explored in order to identify the barriers to implementation as human operators transitionally fulfill roles that will be ultimately tasked to future automation. Throughout the tests, similar issues were regularly documented and are expected to persist if not formally addressed by consistent procedures or regulation. Documented here, along with suggested mitigations, are the most frequently noted HSI items, which include operator training, data standardization, and information quantity.

UTM↗

NASA NTRSDate not supplied

Machine Learning Explainability and Transferability for Path Navigation

Deep neural networks are powerful tools for machine perception. Unfortunately their decisions are difficult to explain due to the complexity and size of the networks. Previously we have alleviated this issue by using the representational portion of a deep neural network and combining it with a k-nearest neighbor (KNN) classifier. Through inspection of the decisions made by the KNN, we can directly see the training data responsible for the decisions, allowing us to determine the quality of the overall decision and the quality of the representational layer of the deep NN. While the technique worked well, it requires tens of thousands of latent vectors to be stored for classification. In addition, it lacks the ability to show how parts of an image influence the classification decision. Here we address these issues by 1) Using a radial basis function network (RBFN) in place of the KNN allowing far fewer images to be used in deployment and 2) Using an auto encoder network for explainability. In addition to these techniques, we examine the effects of transfer learning to determine that results are robust. All results are tested on a domain where an unmanned aerial vehicle (UAV) navigates a forest trail through a single camera.

explainable machine learning↗

NASA NTRS2020

Southern Colorado Disasters: Using NASA Observations to Map Aspen Extent and Recovery Due to Wildfire

Quaking aspen (Populus tremuloides) is an important species for wildlife, watershed health, and ecosystem resilience across its range. Heavy ungulate browsing and factors influenced by a changing climate including seasonal temperature changes and moisture deficit have led to reduced post-fire aspen regeneration rates in southern Colorado. This project partnered with Trinchera Ranch and the Colorado State Forest Service to estimate aspen recovery after the Spring Creek Fire, which ignited in June of 2018. The Southern Colorado Disasters team utilized field measurements and satellite imagery from Landsat Operational Land Imager (OLI), Sentinel-2 MultiSpectral Instrument (MSI), and the Shuttle Radar Topography Mission (SRTM) to train and run several random forest models that detect pre- and post-fire aspen extent. Ocular sampling of over 500 points on high-resolution pre-fire and post-fire images identified percentage aspen cover in 30 x 30-meter grid cells. This process provided training data for regression models, which were able to detect aspen across the landscape for both time periods using multiple remote sensing vegetation health indices. In addition, landscape suitability for aspen regeneration was modeled to provide a guide for managers on where to monitor for aspen regeneration post-fire.

DEVELOP Project Summary↗

NASA NTRS2020

Southern Colorado Disasters: Using NASA Earth Observations to Map Aspen Extent and Recovery Due to Wildfire

Quaking aspen (Populus tremuloides) is an important species for wildlife, watershed health, and ecosystem resilience across its range. Heavy ungulate browsing and factors influenced by a changing climate including seasonal temperature changes and moisture deficit have led to reduced post-fire aspen regeneration rates in southern Colorado. This project partnered with Trinchera Ranch and the Colorado State Forest Service to estimate aspen recovery after the Spring Creek Fire, which ignited in June of 2018. The Southern Colorado Disasters team utilized field measurements and satellite imagery from Landsat Operational Land Imager (OLI), Sentinel-2 MultiSpectral Instrument (MSI), and the Shuttle Radar Topography Mission (SRTM) to train and run several random forest models that detect pre- and post-fire aspen extent. Ocular sampling of over 500 points on high-resolution pre-fire and post-fire images identified percentage aspen cover in 30 x 30-meter grid cells. This process provided training data for regression models, which were able to detect aspen across the landscape for both time periods using multiple remote sensing vegetation health indices. In addition, landscape suitability for aspen regeneration was modeled to provide a guide for managers on where to monitor for aspen regeneration post-fire.

DEVELOP Technical Paper↗

NASA NTRS2018

Online Self-Supervised Long-Range Scene Segmentation for MAVs

Recently, there have been numerous advances in the development of payload and power constrained lightweight Micro Aerial Vehicles (MAVs). As these robots aspire for high-speed autonomous ﬂights in complex dynamic environments, robust scene understanding at long-range becomes critical. The problem is heavily characterized by either the limitations imposed by sensor capabilities for geometry-based methods, or the need for large-amounts of manually annotated training data required by data-driven methods. This motivates the need to build systems that have the capability to alleviate these problems by exploiting the complimentary strengths of both geometry and data-driven methods. In this paper, we take a step in this direction and propose a generic framework for adaptive scene segmentation using self-supervised online learning. We present this in the context of vision-based autonomous MAV ﬂight, and demonstrate the efﬁcacy of our proposed system through extensive experiments on benchmark datasets and real world ﬁeld tests.

Matthies, Larry↗

NASA NTRSDate not supplied

Studying Human-System Interaction within the UTM System

Over five years, NASA, together with partnering organizations, has been developing and successfully demonstrating the maturing capabilities of the Unmanned Aircraft Systems (UAS) Traffic Management (UTM) system and its ability to support communication and coordination among small UAS operations through a series of flight tests. During these flight tests, human-system interaction (HSI) elements were also explored in order to identify the barriers to implementation as human operators transitionally fulfill roles that will be ultimately tasked to future automation. Throughout the tests, similar issues were regularly documented and are expected to persist if not formally addressed by consistent procedures or regulation. Documented here, along with suggested mitigations, are the most frequently noted HSI items, which include operator training, data standardization, and information quantity. .

UTM↗

NASA NTRSDate not supplied

Studying Human-System Interaction with the UTM System

Over five years, NASA, together with partnering organizations, has been developing and successfully demonstrating the maturing capabilities of the Unmanned Aircraft Systems (UAS) Traffic Management (UTM) system and its ability to support communication and coordination among small UAS operations through a series of flight tests. During these flight tests, human-system interaction (HSI) elements were also explored in order to identify the barriers to implementation as human operators transitionally fulfill roles that will be ultimately tasked to future automation. Throughout the tests, similar issues were regularly documented and are expected to persist if not formally addressed by consistent procedures or regulation. Documented here, along with suggested mitigations, are the most frequently noted HSI items, which include operator training, data standardization, and information quantity.

UTM↗

NASA NTRSDate not supplied

Embracing the fourth Industrial Revolution - Challenges, Opportunities and Path Forward for Propulsion

The aerospace industry is at a point where components are reaching design maturity and performance improvements are incremental. Aggressive goals to achieve sustainability and the threat of climate change necessitate a new paradigm. An artificial intelligence (AI) approach that enables revolutionary changes in system architecture, mission analysis and performance metrics is needed. The growing interest and development in the field of machine learning presents an opportunity to speed up by 10X or more the discovery, analysis and development of aerospace systems using artificial intelligence. Through IDEAS (Intelligent Design and Engineering of Aerospace Systems) we are embarking on a research and development effort that addresses this opportunity. The objective of IDEAS is to enable design of systems based on requirements. We discuss possible approaches to generating data, training models and applying them to near term applications. Results of recent workshops with industry, academia and other agencies to identify challenges to adopting AI and machine learning are presented.

Machine learning↗

NASA NTRSDate not supplied

Embracing the Fourth Industrial Revolution - Challenges, Opportunities and Path Forward for Propulsion

The aerospace industry is at a point where components are reaching design maturity and performance improvements are incremental. Aggressive goals to achieve sustainability and the threat of climate change necessitate a new paradigm. An artificial intelligence (AI) approach that enables revolutionary changes in system architecture, mission analysis and performance metrics is needed. The growing interest and development in the field of machine learning presents an opportunity to speed up by 10X or more the discovery, analysis and development of aerospace systems using artificial intelligence. Through IDEAS (Intelligent Design and Engineering of Aerospace Systems) we are embarking on a research and development effort that addresses this opportunity. The objective of IDEAS is to enable design of systems based on requirements. We discuss possible approaches to generating data, training models and applying them to near term applications. Results of recent workshops with industry, academia and other agencies to identify challenges to adopting AI and machine learning are presented.

Machine learning↗

NASA NTRS2021

Machine Learning Based Path Planning for Improved Rover Navigation

Enhanced AutoNav (ENav), the baseline surface navigation software for NASA’s Perseverance rover, sorts a list of candidate paths for the rover to traverse, then uses the Approximate Clearance Evaluation (ACE) algorithm to evaluate whether the most highly ranked paths are safe. ACE is crucial for maintaining the safety of the rover, but is computationally expensive. If the most promising candidates in the list of paths are all found to be infeasible, ENav must continue to search the list and run time-consuming ACE evaluations until a feasible path is found. In this paper, we present two heuristics that, given a terrain heightmap around the rover, produce cost estimates that more effectively rank the candidate paths before ACE evaluation. The first heuristic uses Sobel operators and convolution to incorporate the cost of traversing high-gradient terrain. The second heuristic uses a machine learning (ML) model to predict areas that will be deemed untraversable by ACE. We used physics simulations to collect training data for the ML model and to run Monte Carlo trials to quantify navigation performance across a variety of terrains with various slopes and rock distributions. Compared to ENav's baseline performance, integrating the heuristics can lead to a significant reduction in ACE evaluations and average computation time per planning cycle, increase path efficiency, and maintain or improve the rate of successful traverses. This strategy of targeting specific bottlenecks with ML while maintaining the original ACE safety checks provides an example of how ML can be infused into planetary science missions and other safety-critical software.

Yue, Yisong↗

NASA NTRSDate not supplied

A Reference Ocean Surface Emission and Backscatter Model from Microwaves to Infrared

Satellite observations are vital for the initialization of Numerical Weather Prediction models, and very important for climate monitoring and prediction, as well as other applications such as hydrology and flood awareness prediction. Knowledge of radiative contributions from the Earth's surface is needed to sound the lower troposphere from space. The lack of a reference quality ocean emission and backscatter model is a major gap in our ability to provide absolute calibration of the satellite based observing system. Uncertainty in emissivity models is not well characterized and different models are used for different spectral bands, for active and passive instruments. An International Space Science Institute (ISSI) team was put together [4] to address these issues. The objectives of the team are to provide a reference model as a community software (i.e., documented and freely available code), that is maintained and supported, has traceable uncertainty estimations, and that enables new science from microwaves to infrared with bidirectional reflectance distribution function (BRDF) capability. We will present the model and its various components, discussing the choices between various parameterizations, building on the LOCEAN model of [2]. The model predictions will be evaluated at various frequencies, including comparisons to radiometric observations by SMAP, AMSR2 and GMI (e.g., [5]). We will discuss early model evaluation in the infrared and for active microwave sensors. Areas of ongoing research include improving the foam parametrization (coverage and emissivity) to provide consistent performances across frequencies, building on [1], and the azimuthal dependence of the active and passive signals. The model will be used to generate training data for fast models e.g., Fastem, [3], that are used in operational data assimilation and climate re-analysis.

Emmanuel Dinnat↗