Search NASA⌕ Search

SEARCH · Search NASA

Results for “Generative 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 487 records · Page 27

Multiresolutional models of uncertainty generation and reduction

Kolmogorov's axiomatic principles of the probability theory, are reconsidered in the scope of their applicability to the processes of knowledge acquisition and interpretation. The model of uncertainty generation is modified in order to reflect the reality of engineering problems, particularly in the area of intelligent control. This model implies algorithms of learning which are organized in three groups which reflect the degree of conceptualization of the knowledge the system is dealing with. It is essential that these algorithms are motivated by and consistent with the multiresolutional model of knowledge representation which is reflected in the structure of models and the algorithms of learning.

Meystel, A.↗

Space Station Freedom Gateway to the Future

The first inhabited outpost on the frontier of space will be a place to live, work, and discover. Experiments conducted on Freedom will advance scientific knowledge about our world, our environment, and ourselves. We will learn how to adapt to the space environment and to build and operate new spacecraft with destinations far beyond Earth, continuing the tradition of exploration that began with a journey to the Moon. What we learn from living and working on Freedom will strengthen our expertise in science and engineering, promote national research and development initiatives and inspire another generation of Americans to push forward and onward. On the eve of the 21st century, Space Station Freedom will be our gateway to the future. This material covers gateways to space, research, discovery, utilization, benefits, and NASA.

Source record↗

Lessons from Adaptive Level One Accelerator (ALOA) System Implementation

The Adaptive Level One Accelerator (ALOA) system was developed as part of the Earth Science Data and Information System (ESDIS) project. The reconfigurable computing technologies were investigated for Level 1 satellite telemetry data processing to achieve computing acceleration and cost reduction for the next-generation Level 1 data processing systems. The MODIS instrument calibration algorithm was implemented using reconfigurable a computer. The system development process and the lessons learned throughout the design cycle are summarized in this paper.

Patel, Umesh D.↗

Intelligent Contingency Management for Urban Air Mobility

The third aviation revolution is seeking to enable transportation where users have access to immediate and flexible air travel; the users dictate trip origin, destination and timing. One of the major components of this vision is urban air mobility (UAM) for the masses. UAM means a safe and efficient system for vehicles to move passengers and cargo within a city. In order to reach UAM’s full market potential the vehicle will have to be autonomous. One of the primary challenges of autonomous flight is dealing with off-nominal events, both common and unforeseen; thus, intelligent contingency management (ICM) is one of the enabling technologies. In this context, the vehicle has to be aware of its internal state and external environment at all times, ascertain its capability and make decisions about mission completion or modification. All of these functions require data to model and assess the environment and then take actions based on these models. Necessarily, there is uncertainty associated with the data and the models generated from it. Since we are dealing with safety-critical systems, one of the main challenges of ICM is to generate sufficient data and to minimize its uncertainty to enable practical and safe decision making. We propose an overall architecture that incorporates deterministic and learning algorithms together to assess vehicle capabilities, project these into the future and make decisions on mission management level. A layered approach allows for mature parts and technologies to be integrated into early highly automated vehicles before the final state of autonomy is reached.

data-driven systems↗

Developing Open-Source Training Materials for AI/ML and Space Biological Sciences Using NASA Cloud-Based Data

Artificial Intelligence (AI) and Machine Learning (ML) has gained significant traction in the biological and biomedical research fields, in part due to a culture of open data sharing and reuse. AI/ML methodology is well-suited to recognize and predict biological patterns from high-dimensional next-generation sequencing data (e.g. whole genome sequencing, transcriptomic sequencing), as well as from biological or medical imaging data (e.g. microscopy, computed tomography, ultrasound, magnetic resonance imaging, radiography). These methodologies hold particular promise for space biosciences research and automated space health monitoring systems. However, there are key considerations for properly training, validating, and testing a machine learning model in biological research or clinical application. Inexperienced researchers can produce models that perform poorly outside of the training dataset. Open Science principles such as data sharing and open-source code must go hand-in-hand with publicly available, high-quality training curricula in best practices, with modules centered on real-life scientific use cases and data so future AI/ML practitioners gain experience on real problems. Here we present the development of open-source training materials for AI/ML and space biosciences, as part of the NASA Transform to Open Science Training (TOPST) initiative. We develop 4 independent training programs, focused on the following topics: 1) Fundamentals of Machine Learning and Space Biosciences Domain, 2) Open Science, Artificial Intelligence, and Ethical Best Practices for Data Sharing and Analysis, 3) Using AI/ML Classification to Identify Gene Networks Affected By Space Exposure in Mouse Liver, and 4) Using Neural Networks to Find DNA Damage Patterns in Immune Cells after Radiation. All programs leverage cloud-based NASA biological datasets. The curriculum we present will enable worldwide access to training in AI/ML and scientific analysis.

James Casaletto↗

MODIS On-Orbit Calibration and Lessons Learned

The Moderate Resolution Imaging Spectroradiometer (MODIS) is a key instrument for NASA's Earth Observing System (EOS) Terra and Aqua missions. Since launch, Terra and Aqua MODIS have successfully operated for more than 12 and 10 years, respectively, and generated an unprecedented amount of data products for the science and user community over a wide range of applications. MODIS was developed with improved design and stringent calibration requirements over its heritage sensors in order . to extend and enhance their long-term data records. Its follow-on instrument, the Visible/Infrared Imager Radiometer Suite (VIIRS), was launched on-board the Suomi National Polar-orbiting Partnership (NPP) spacecraft October 28, 2011. MODIS collects data in 36 spectral bands, covering wavelengths from 0.41 to 14.S!Jlll, and at 250m, SOOm, and lkm spatial resolutions (nadir). MODIS on-orbit calibration is provided by a set of onboard calibrators (OBC), including a solar diffuser (SO), a solar diffuser stability monitor (SDSM), a blackbody (BB), and a spectroradiometric calibration assembly (SRCA). In addition to the onboard calibrators, regular lunar observations are made by both Terra and Aqua MODIS to track their calibration stability in the reflective solar region. This tutorial session provides an overview of MODIS on-orbit calibration and characterization methodologies. It discusses challenging issues and lessons learned from sensor design, operation, calibration, and inter-comparisons. Examples of instrument on-orbit performance are illustrated with a focus on the improvements made based on various lessons learned. It is expected that MODIS experience and lessons will continue to provide valuable information for future earth observing missions/sensors.

Xiong, Jack↗

Architecting Safer Autonomous Aviation Systems

The aviation literature gives relatively little guidance to practitioners about the specifics of architecting systems for safety, particularly the impact of architecture on allocating safety requirements, or the relative ease of system assurance resulting from system or subsystem level architectural choices. As an exemplar, this paper considers common architectural patterns used within traditional aviation systems and explores their safety and safety assurance implications when applied in the context of integrating artificial intelligence (AI) and machine learning (ML) based functionality. Considering safety as an architectural property, we discuss both the allocation of safety requirements and the architectural trade-offs involved early in the design lifecycle. This approach could be extended to other assured properties, similar to safety, such as security. We conclude with a discussion of the safety considerations that emerge in the context of candidate architectural patterns that have been proposed in the recent literature for enabling autonomy capabilities by integrating AI and ML. A recommendation is made for the generation of a property-driven architectural pattern catalogue.

Architecture patterns↗

Onboard Nonlinear Engine Sensor and Component Fault Diagnosis and Isolation Scheme

A method detects and isolates in-flight sensor, actuator, and component faults for advanced propulsion systems. In sharp contrast to many conventional methods, which deal with either sensor fault or component fault, but not both, this method considers sensor fault, actuator fault, and component fault under one systemic and unified framework. The proposed solution consists of two main components: a bank of real-time, nonlinear adaptive fault diagnostic estimators for residual generation, and a residual evaluation module that includes adaptive thresholds and a Transferable Belief Model (TBM)-based residual evaluation scheme. By employing a nonlinear adaptive learning architecture, the developed approach is capable of directly dealing with nonlinear engine models and nonlinear faults without the need of linearization. Software modules have been developed and evaluated with the NASA C-MAPSS engine model. Several typical engine-fault modes, including a subset of sensor/actuator/components faults, were tested with a mild transient operation scenario. The simulation results demonstrated that the algorithm was able to successfully detect and isolate all simulated faults as long as the fault magnitudes were larger than the minimum detectable/isolable sizes, and no misdiagnosis occurred

Tang, Liang↗

Software Template for Instruction in Mathematics

Intelligent Math Tutor (IMT) is a software system that serves as a template for creating software for teaching mathematics. IMT can be easily connected to artificial-intelligence software and other analysis software through input and output of files. IMT provides an easy-to-use interface for generating courses that include tests that contain both multiple-choice and fill-in-the-blank questions, and enables tracking of test scores. IMT makes it easy to generate software for Web-based courses or to manufacture compact disks containing executable course software. IMT also can function as a Web-based application program, with features that run quickly on the Web, while retaining the intelligence of a high-level language application program with many graphics. IMT can be used to write application programs in text, graphics, and/or sound, so that the programs can be tailored to the needs of most handicapped persons. The course software generated by IMT follows a "back to basics" approach of teaching mathematics by inducing the student to apply creative mathematical techniques in the process of learning. Students are thereby made to discover mathematical fundamentals and thereby come to understand mathematics more deeply than they could through simple memorization.

Shelton, Robert O.↗

Program Facilitates CMMI Appraisals

A computer program has been written to facilitate appraisals according to the methodology of Capability Maturity Model Integration (CMMI). [CMMI is a government/industry standard, maintained by the Software Engineering Institute at Carnegie Mellon University, for objectively assessing the engineering capability and maturity of an organization (especially, an organization that produces software)]. The program assists in preparation for a CMMI appraisal by providing drop-down lists suggesting required artifacts or evidence. It identifies process areas for which similar evidence is required and includes a copy feature that reduces or eliminates repetitive data entry. It generates reports to show the entire framework for reference, the appraisal artifacts to determine readiness for an appraisal, and lists of interviewees and questions to ask them during the appraisal. During an appraisal, the program provides screens for entering observations and ratings, and reviewing evidence provided thus far. Findings concerning strengths and weaknesses can be exported for use in a report or a graphical presentation. The program generates a chart showing capability level ratings of the organization. A context-sensitive Windows help system enables a novice to use the program and learn about the CMMI appraisal process.

Sweetser, Wesley↗

Cognitive Communications and Networking Technology Infusion Study Report

As the envisioned next-generation SCaN Network transitions into an end-to-end “system of systems” with new enabling capabilities, it is anticipated that the introduction of machine learning, artificial intelligence, and other cognitive strategies into the network infrastructure will result in increased mission science return, improved resource efficiencies, and increased autonomy and reliability. This enhanced set of cognitive capabilities will be implemented via a “space cloud” concept to achieve a service-oriented architecture with distributed cognition, de-centralized routing, and shared, on-orbit data processing. The enabling cognitive communications and networking capabilities that may facilitate the desired network enhancements are identified in this document, and the associated enablers of these capabilities, such as technologies and standards, are described in detail.

Knoblock, Eric J.↗

Creating Formal Characterizations of Routine Contingency Management in Commercial Aviation

The identification, modelling, and analysis of root causes of accidents and incidents dominate conventional safety management approaches. However, the effect of humans’ safety-producing behavior on the overall resilience of the system is often neglected. Additionally, emerging aviation markets are giving rise to concepts of operation, such as urban air mobility and optionally piloted air cargo operations, that are leading to a shift in locus of control between humans and automation. Without an understanding of the human contribution to safety, it is difficult to assess the effects of these novel role allocations on overall system safety. In this work, safety-producing behaviors are identified and abstracted into resilient performance strategies. Production rules that encapsulate these strategies are then generated and classified in the Soar cognitive architecture. The strategies are then applied to a remotely-operated air cargo example to demonstrate how safe learning is facilitated. The learned rules and strategies are then formally verified.

Safety Critical Systems↗

Implementation of Machine Learning Methods for Crater-Based Navigation

Terrain Relative Navigation methods require surface feature detectors to gain information from images used to improve on-board state estimates. This paper presents the development of a crater detection method based on Machine Learning that can extract data from optical images with different crater shapes and sizes, under varying lighting conditions. This work includes an automated capability for generating labeled training data and iterative testing of the neural network-based crater detector. Preliminary results are included to quantify the detector’s accuracy compared to a known crater catalog, given a set of real lunar images from the Lunar Reconnaissance Orbiter.

Sofia G Catalan↗

Coevolution of Machine Learning and Process-Based Modelling to Revolutionize Earth and Environmental Sciences: A Perspective

Machine learning (ML) applications in Earth and environmental sciences (EES) have gained incredible momentum in recent years. However, these ML applications have largely evolved in ‘isolation’ from the mechanistic, process-based modelling (PBM) paradigms, which have historically been the cornerstone of scientific discovery and policy support. In this perspective, we assert that the cultural barriers between the ML and PBM communities limit the potential of ML, and even its ‘hybridization’ with PBM, for EES applications. Fundamental, but often ignored, differences between ML and PBM are discussed as well as their strengths and weaknesses in light of three overarching modelling objectives in EES, (1) nowcasting and prediction, (2) scenario analysis, and (3) diagnostic learning. The paper ponders over a ‘coevolutionary’ approach to model building, shifting away from a borrowing to a co-creation culture, to develop a generation of models that leverage the unique strengths of ML such as scalability to big data and high-dimensional mapping, while remaining faithful to process-based knowledge base and principles of model explainability and interpretability, and therefore, falsifiability.

Saman Razavi↗

AstroDOME: Standardized, Mass-Produceable, Small NASA Astrophysics Observatories for Big Missions

NASA’s Goddard Space Flight Center is investigating the use of small, commercial off the shelf (COTS) spacecraft to enable breakthrough astrophysics research at radically reduced cost over previous generations of missions. This paper will focus on the Mission Systems perspective of this investigation. It will cover the spacecraft to instrument interface, and lessons learned in early concept payload development. Low-cost COTS spacecraft have proliferated in the last ~5 years. They represent an order of magnitude reduction in the cost of spacecraft bus development and acquisition. They are enabled by the advent of low-cost smallsat launch options in the immediately preceding years, and are seeded by recent distributed and disaggregated missions in other sectors such as telecommunications megaconstellations (e.g. Starlink, OneWeb, Project Kuiper) and ongoing Space Development Agency (SDA) developments. Our study was tasked with developing a standardized, mass-produceable, astrophysics payload which is reconfigurable to different wavelength bands while maintaining a single optics bench design, and a standardized mechanical, electrical, data, and thermal interface to the spacecraft bus, and which may be reconfigured to multiple COTS spacecraft offerings. In this paper we will discuss lessons learned from our case study. We will also present identified keys to mission success for performing cutting edge astrophysics observations while taking advantage of economies of scale with a disaggregated, standardized architecture.

Matthew Marcus↗

Advancing Wildfire Monitoring with TEMPO and ML tools: Hourly Smoke and Fire‑Front Mapping and Near‑Surface NO₂ Predictions

Wildfires impose substantial impacts on communities and regions downwind of wildfire smoke. We present a TEMPO‑enabled workflow that generates value‑added Level‑3 smoke‑plume masks and fire‑front maps for large wildfires, such as 2024 Park Fire, using the self‑supervised deep learning system SIT‑FUSE, along with near‑surface NO₂ predictions produced by a foundation model (Microsoft Aurora). We conclude by outlining a roadmap for expanding these capabilities to additional Western U.S. wildfire events and for delivering actionable tools to stakeholders. This open-source, reproducible workflow provides a scalable framework for cross-agency wildfire monitoring to overcome traditional limitations in smoke-cloud discrimination and air-quality forecasting by incorporating TEMPO data and beyond.

Xiaohua Pan↗

Space Vehicle Powerdown Philosophies Derived from the Space Shuttle Program

In spaceflight, electrical power is a vital but limited resource. Almost every spacecraft system, from avionics to life support systems, relies on electrical power. Since power can be limited by the generation system s performance, available consumables, solar array shading, or heat rejection capability, vehicle power management is a critical consideration in spacecraft design, mission planning, and real-time operations. The purpose of this paper is to capture the powerdown philosophies used during the Space Shuttle Program. This paper will discuss how electrical equipment is managed real-time to adjust the overall vehicle power level to ensure that systems and consumables will support changing mission objectives, as well as how electrical equipment is managed following system anomalies. We will focus on the power related impacts of anomalies in the generation systems, air and liquid cooling systems, and significant environmental events such as a fire, decrease in cabin pressure, or micrometeoroid debris strike. Additionally, considerations for executing powerdowns by crew action or by ground commands from Mission Control will be presented. General lessons learned from nearly 30 years of Space Shuttle powerdowns will be discussed, including an in depth case-study of STS-117. During this International Space Station (ISS) assembly mission, a failure of computers controlling the ISS guidance, navigation, and control system required that the Space Shuttle s maneuvering system be used to maintain attitude control. A powerdown was performed to save power generation consumables, thus extending the docked mission duration and allowing more time to resolve the issue.

Willsey, Mark↗

Machine Learning Approach for Aircraft Performance Model Parameter Estimation for Trajectory Prediction Applications

Inaccurate prediction of aircraft trajectory by ground-based decision support tools (DST) is a major concern in air traffic management (ATM). Aircraft trajectory prediction tools rely on a simplified point-mass aircraft performance model (APM) to make their predictions. Even though the performance coefficients and weight of an aircraft are a vital part of the APM’s predictions and accuracy, these coefficients are proprietary in nature and therefore, unavailable to DSTs. Current ATM research focuses on improving the estimate of some APM parameters by freezing all other coefficients. This simplified approach introduces unwanted sources of bias and negatively impacts the accuracy of the performance model. In this paper, we apply machine learning (ML) techniques for the simultaneous prediction of three key APM parameters (two drag coefficients and the initial aircraft weight). To accomplish this, we employ an ordinary differential equation (ODE) fitting approach to generate optimized APM parameter labels customized to each individual flight record. Subsequently, we train ML models to capture the relationship between the historical data and the optimized APM parameters. Two different ML model solutions are applied and APM coefficients are predicted for unseen flights. The results indicate that the ML models are able to capture the relationship between APM parameters and flight-related features with good accuracy.

trajectory prediction, machine learning, aircraft ↗