Search NASA⌕ Search

SEARCH · Search NASA

Results for “Domain knowledge”

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 253 records · Page 14

An Experiment in Scientific Program Understanding

This paper concerns a procedure that analyzes aspects of the meaning or semantics of scientific and engineering code. This procedure involves taking a user's existing code, adding semantic declarations for some primitive variables, and parsing this annotated code using multiple, independent expert parsers. These semantic parsers encode domain knowledge and recognize formulae in different disciplines including physics, numerical methods, mathematics, and geometry. The parsers will automatically recognize and document some static, semantic concepts and help locate some program semantic errors. Results are shown for three intensively studied codes and seven blind test cases; all test cases are state of the art scientific codes. These techniques may apply to a wider range of scientific codes. If so, the techniques could reduce the time, risk, and effort required to develop and modify scientific codes.

Stewart, Mark E. M.↗

Synthesizing Certified Code

Code certification is a lightweight approach to demonstrate software quality on a formal level. Its basic idea is to require producers to provide formal proofs that their code satisfies certain quality properties. These proofs serve as certificates which can be checked independently. Since code certification uses the same underlying technology as program verification, it also requires many detailed annotations (e.g., loop invariants) to make the proofs possible. However, manually adding theses annotations to the code is time-consuming and error-prone. We address this problem by combining code certification with automatic program synthesis. We propose an approach to generate simultaneously, from a high-level specification, code and all annotations required to certify generated code. Here, we describe a certification extension of AUTOBAYES, a synthesis tool which automatically generates complex data analysis programs from compact specifications. AUTOBAYES contains sufficient high-level domain knowledge to generate detailed annotations. This allows us to use a general-purpose verification condition generator to produce a set of proof obligations in first-order logic. The obligations are then discharged using the automated theorem E-SETHEO. We demonstrate our approach by certifying operator safety for a generated iterative data classification program without manual annotation of the code.

Whalen, Michael↗

Extending a Flight Management Computer for Simulation and Flight Experiments

In modern transport aircraft, the flight management computer (FMC) has evolved from a flight planning aid to an important hub for pilot information and origin-to-destination optimization of flight performance. Current trends indicate increasing roles of the FMC in aviation safety, aviation security, increasing airport capacity, and improving environmental impact from aircraft. Related research conducted at the Langley Research Center (LaRC) often requires functional extension of a modern, full-featured FMC. Ideally, transport simulations would include an FMC simulation that could be tailored and extended for experiments. However, due to the complexity of a modern FMC, a large investment (millions of dollars over several years) and scarce domain knowledge are needed to create such a simulation for transport aircraft. As an intermediate alternative, the Flight Research Services Directorate (FRSD) at LaRC created a set of reusable software products to extend flight management functionality upstream of a Boeing-757 FMC, transparently simulating or sharing its operator interfaces. The paper details the design of these products and highlights their use on NASA projects.

Madden, Michael M.↗

Revealing the ISO/IEC 9126-1 Clique Tree for COTS Software Evaluation

Previous research has shown that acyclic dependency models, if they exist, can be extracted from software quality standards and that these models can be used to assess software safety and product quality. In the case of commercial off-the-shelf (COTS) software, the extracted dependency model can be used in a probabilistic Bayesian network context for COTS software evaluation. Furthermore, while experts typically employ Bayesian networks to encode domain knowledge, secondary structures (clique trees) from Bayesian network graphs can be used to determine the probabilistic distribution of any software variable (attribute) using any clique that contains that variable. Secondary structures, therefore, provide insight into the fundamental nature of graphical networks. This paper will apply secondary structure calculations to reveal the clique tree of the acyclic dependency model extracted from the ISO/IEC 9126-1 software quality standard. Suggestions will be provided to describe how the clique tree may be exploited to aid efficient transformation of an evaluation model.

Morris, A. Terry↗

Peer-to-Peer Planning for Space Mission Control

Planning and scheduling for space operations entails the development of applications that embed intimate domain knowledge of distinct areas of mission control, while allowing for significant collaboration among them. The separation is useful because of differences in the planning problem, solution methods, and frequencies of replanning that arise in the different disciplines. For example, planning the activities of human spaceflight crews requires some reasoning about all spacecraft resources at timescales of minutes or seconds, and is subject to considerable volatility. Detailed power planning requires managing the complex interplay of power consumption and production, involves very different classes of constraints and preferences, but once plans are generated they are relatively stable.

Barreiro, Javier↗

Testing First-Order Logic Axioms in AutoCert

AutoCert [2] is a formal verification tool for machine generated code in safety critical domains, such as aerospace control code generated from MathWorks Real-Time Workshop. AutoCert uses Automated Theorem Provers (ATPs) [5] based on First-Order Logic (FOL) to formally verify safety and functional correctness properties of the code. These ATPs try to build proofs based on user provided domain-specific axioms, which can be arbitrary First-Order Formulas (FOFs). These axioms are the most crucial part of the trusted base, since proofs can be submitted to a proof checker removing the need to trust the prover and AutoCert itself plays the part of checking the code generator. However, formulating axioms correctly (i.e. precisely as the user had really intended) is non-trivial in practice. The challenge of axiomatization arise from several dimensions. First, the domain knowledge has its own complexity. AutoCert has been used to verify mathematical requirements on navigation software that carries out various geometric coordinate transformations involving matrices and quaternions. Axiomatic theories for such constructs are complex enough that mistakes are not uncommon. Second, adjusting axioms for ATPs can add even more complexity. The axioms frequently need to be modified in order to have them in a form suitable for use with ATPs. Such modifications tend to obscure the axioms further. Thirdly, speculating validity of the axioms from the output of existing ATPs is very hard since theorem provers typically do not give any examples or counterexamples.

Ahn, Ki Yung↗

Multiple Damage Progression Paths in Model-Based Prognostics

Model-based prognostics approaches employ domain knowledge about a system, its components, and how they fail through the use of physics-based models. Component wear is driven by several different degradation phenomena, each resulting in their own damage progression path, overlapping to contribute to the overall degradation of the component. We develop a model-based prognostics methodology using particle filters, in which the problem of characterizing multiple damage progression paths is cast as a joint state-parameter estimation problem. The estimate is represented as a probability distribution, allowing the prediction of end of life and remaining useful life within a probabilistic framework that supports uncertainty management. We also develop a novel variance control mechanism that maintains an uncertainty bound around the hidden parameters to limit the amount of estimation uncertainty and, consequently, reduce prediction uncertainty. We construct a detailed physics-based model of a centrifugal pump, to which we apply our model-based prognostics algorithms. We illustrate the operation of the prognostic solution with a number of simulation-based experiments and demonstrate the performance of the chosen approach when multiple damage mechanisms are active

Daigle, Matthew↗

A Model-Based Prognostics Approach Applied to Pneumatic Valves

Within the area of systems health management, the task of prognostics centers on predicting when components will fail. Model-based prognostics exploits domain knowledge of the system, its components, and how they fail by casting the underlying physical phenomena in a physics-based model that is derived from first principles. Uncertainty cannot be avoided in prediction, therefore, algorithms are employed that help in managing these uncertainties. The particle filtering algorithm has become a popular choice for model-based prognostics due to its wide applicability, ease of implementation, and support for uncertainty management. We develop a general model-based prognostics methodology within a robust probabilistic framework using particle filters. As a case study, we consider a pneumatic valve from the Space Shuttle cryogenic refueling system. We develop a detailed physics-based model of the pneumatic valve, and perform comprehensive simulation experiments to illustrate our prognostics approach and evaluate its effectiveness and robustness. The approach is demonstrated using historical pneumatic valve data from the refueling system.

Daigle, Matthew J.↗

Semi-Supervised Eigenbasis Novelty Detection

Recent discoveries in high-time-resolution radio astronomy data have focused attention on a new class of events. Fast transients are rare pulses of radio frequency energy lasting from microseconds to seconds that might be produced by a variety of exotic astrophysical phenomena. For example, X-ray bursts, neutron stars, and active galactic nuclei are all possible sources of short-duration, transient radio signals. It is difficult to anticipate where such signals might appear, and they are most commonly discovered through analysis of high-time- resolution data that had been collected for other purposes. Transients are often faint and difficult to detect, so improved detection algorithms can directly benefit the science yield of all such commensal monitoring. A new detection algorithm learns a low-dimensional linear manifold for describing the normal data. High reconstruction error indicates a novel signal that does not match the patterns of normal data. One unsupervised portion of the manifold model adapts its representation in response to recent data. A second supervised portion of the model is made of a basis trained in advance using labeled examples of RFI; this prevents false positives due to these events. For a linear model, an orthonormalization operation is used to combine these bases prior to the anomaly detection decision. Another novel aspect of the approach lies in combining basis vectors learned in an unsupervised, online fashion from the data stream with supervised basis vectors learned in advance from known examples of false alarms. Adaptive, data-driven detection is achieved that is also informed by existing domain knowledge about signals that may be statistically anomalous, but are not interesting and should therefore be ignored. The method was evaluated using data from the Parkes Multibeam Survey. This data set was originally collected to search for pulsars, which are astronomical sources that emit radio pulses at regular periods. However, several non-pulsar anomalies have recently been discovered in this dataset, making it a compelling test case. By explicitly filtering known false alarm patterns, the approach yields significantly better performance than current transient detection methods.

Wagstaff, Kiri L.↗

A Comparison of Filter-based Approaches for Model-based Prognostics

Model-based prognostics approaches use domain knowledge about a system and its failure modes through the use of physics-based models. Model-based prognosis is generally divided into two sequential problems: a joint state-parameter estimation problem, in which, using the model, the health of a system or component is determined based on the observations; and a prediction problem, in which, using the model, the stateparameter distribution is simulated forward in time to compute end of life and remaining useful life. The first problem is typically solved through the use of a state observer, or filter. The choice of filter depends on the assumptions that may be made about the system, and on the desired algorithm performance. In this paper, we review three separate filters for the solution to the first problem: the Daum filter, an exact nonlinear filter; the unscented Kalman filter, which approximates nonlinearities through the use of a deterministic sampling method known as the unscented transform; and the particle filter, which approximates the state distribution using a finite set of discrete, weighted samples, called particles. Using a centrifugal pump as a case study, we conduct a number of simulation-based experiments investigating the performance of the different algorithms as applied to prognostics.

prognostics↗

The Challenge of Grounding Planning in Simulation with an Interactive Model Development Environment

A principal obstacle to fielding automated planning systems is the difficulty of modeling. Physical systems are modeled conventionally based on specification documents and the modeler's understanding of the system. Thus, the model is developed in a way that is disconnected from the system's actual behavior and is vulnerable to manual error. Another obstacle to fielding planners is testing and validation. For a space mission, generated plans must be validated often by translating them into command sequences that are run in a simulation testbed. Testing in this way is complex and onerous because of the large number of possible plans and states of the spacecraft. Though, if used as a source of domain knowledge, the simulator can ease validation. This paper poses a challenge: to ground planning models in the system physics represented by simulation. A proposed, interactive model development environment illustrates the integration of planning and simulation to meet the challenge. This integration reveals research paths for automated model construction and validation.

radio occultation↗

A Generalized Timeline Representation, Services, and Interface for Automating Space Mission Operations

Numerous automated and semi-automated planning & scheduling systems have been developed for space applications. Most of these systems are model-based in that they encode domain knowledge necessary to predict spacecraft state and resources based on initial conditions and a proposed activity plan. The spacecraft state and resources as often modeled as a series of timelines, with a timeline or set of timelines to represent a state or resource key in the operations of the spacecraft. In this paper, we first describe a basic timeline representation that can represent a set of state, resource, timing, and transition constraints. We describe a number of planning and scheduling systems designed for space applications (and in many cases deployed for use of ongoing missions) and describe how they do and do not map onto this timeline model.

Chien, Steve A.↗

Module Architecture for in Situ Space Laboratories

The paper analyzes internal outfitting architectures for space exploration laboratory modules. ISS laboratory architecture is examined as a baseline for comparison; applicable insights are derived. Laboratory functional programs are defined for seven planet-surface knowledge domains. Necessary and value-added departures from the ISS architecture standard are defined, and three sectional interior architecture options are assessed for practicality and potential performance. Contemporary guidelines for terrestrial analytical laboratory design are found to be applicable to the in-space functional program. Densepacked racks of system equipment, and high module volume packing ratios, should not be assumed as the default solution for exploration laboratories whose primary activities include un-scriptable investigations and experimentation on the system equipment itself.

planetary↗

Predicting Real-Time Safety of the National Airspace System

To gain the situational awareness necessary for informed decision making regarding avoidance of airspace hazards, each operator must consolidate operations-relevant information from disparate sources and apply extensive domain knowledge to correctly interpret not just the current state of the NAS but forecast its (combined) evolution over the duration of the operation. This time- and workload-intensive process is periodically repeated throughout the operation so that changes can be managed in a timely manner.The imprecision, inaccuracies, inconsistency, and incompleteness of the incoming data further challenges the process. To facilitate informed decision making, this paper presents a model-based framework for the textitautomated real-time monitoring and prediction of possible effects of airspace hazards on the safety of the National Airspace System (NAS). First, hazards to flight are identified and transformed into sms, that is, quantities of interest that could be evaluated based on available data and are predictive of an unsafe event. The sms and associated thresholds that specify when an event transitions from emphsafe to emphunsafe are combined with models of airspace operations and aircraft dynamics. The framework can include any hazard to flight that can be modeled quantitatively. Models can be detailed and complex, or they can be considerably simplifed, as appropriate to the application. Real-time NAS safety monitoring and prediction begins with an estimate of the state of the NAS using the dynamic models. Given the state estimate and a probability distribution of future inputs to the NAS, we can then predict the evolution of the NAS - the future state - and the occurrence of hazards and unsafe events. The entire probability distribution of airspace sms is computed, not just point estimates, without significant assumptions regarding the distribution type andor parameters. We demonstrate our overall approach through a simulated scenario in which we predict the occurrence of some unsafe events and show how these predictions evolve in time as flight operations progress. Predictions accounting for common sources of uncertainty are included and it is shown how the predictions improve in time, become more confident, and change dynamically as new information is made available to the prediction algorithm.

safety↗

NASA Platform for Autonomous Systems (NPAS)

NASA Platform for Autonomous Systems (NPAS) is a disruptive software platform and processes being developed by the NASA Stennis Space Center (SSC) Autonomous Systems Laboratory (ASL). Autonomous operations are critical for the success, safety and crew survival of NASA deep space missions beyond low Earth orbit, including the Gateway, and for the future of cost-effective ground mission operations. NPAS represents the embodiment of an innovative paradigm for “thinking” autonomy in contrast to brute-force autonomy. NPAS uniquely addresses the requirements and integrates the primary functionalities for autonomous operations, in one platform that includes: (1) Integrated System Health Management (ISHM); (2) autonomy strategies, guided by system health and concepts of operations; (3) domain objects (system elements) and infrastructure to create complete application domain knowledge models (4) infrastructure to create, schedule, and execute mission plans; (5) infrastructure to develop user interfaces for comprehensive awareness; and (6) infrastructure to integrate distributed autonomous applications across networks. NPAS is a single platform that can be used to make any system operate with any desirable degree of autonomy, as well as provide comprehensive system awareness to operators and users.

Figueroa, Fernando↗

Understanding Machine Learning in Earth Science: A Natural Language Processing Approach

Machine learning (ML) is being increasingly utilized in Earth science research. Benefits of ML include efficiency, reduction of human error, and ability to extract hidden patterns within data. However, the mutual lack of each other’s domain knowledge by ML and Earth science stands as a barrier to timely and effective implementation. Earth science, in particular, faces challenges in generating sample data, compared to those of traditional ML problems such as face recognition or stock predictions, where data is abundant and not lacking in ground truth, which is necessary for labeling. Earth science data are more varying in formats, such as HDF5 and image resolutions, and are not standardized across instruments, even within a given Earth science discipline. Previous studies have been done to outline the specific challenges that Earth science faces with ML, while others have focused on using existing publications to mine information efficiently. Other resources such as Scikit-Learn have developed decision trees for choosing appropriate machine learning algorithms, but application within Earth science subjects becomes much more complex. For the current study, we propose a methodology and tool that aids in implementation of ML in Earth science using natural language processing (NLP). Our work comprises three main parts: (1) analyzing existing publications related to ML and Earth science, using natural language processing: (2) extracting from the publications information on ML models subjects in Earth Science: and (3) visualizing the extracted relationships as a network graph. The resulting network graph should aid the Earth science communities in applying optimal ML algorithms and guiding data preparation through visualization of similar studies. The network graph and analysis of document similarity will be the basis of our next step, which is to develop a decision tree for selecting optimal machine learning methodologies for specified Earth science applications.

Zheng, Laura↗

Adaptive Stress Testing of Collision Avoidance Systems for Small UASs with Deep Reinforcement Learning

The next-generation Airborne Collision Avoidance System for smaller UASs (ACAS sXu) is currently being developed and tested by the Federal Aviation Administration (FAA) to provide detect-and-avoid capability for small unmanned aircraft operating beyond line-of-sight. Due to the complexity and safety-critical nature of the system, safety validation is important not only for the certification of the final system, but also for informing changes during the iterative development process. In this paper, we analyze a prototype of ACAS sXu in simulated aircraft encounters to discover scenarios of small near mid-air collisions (sNMACs), an important safety event in which two aircraft come closer than 50 feet horizontally and 15 feet vertically. Due to the size and complexity of the system as well as rarity of sNMAC events, traditional methods such as Monte Carlo testing often require informed setup and targeting to elicit failures. However, such a dependence on domain knowledge can be incompatible with the independent verification and validation (IV&V) process, the aim of which is to discover unforeseen issues. To address these challenges, we apply an accelerated validation method called adaptive stress testing (AST) to find the most likely sNMAC scenarios without reliance on system introspection. AST uses reinforcement learning to adapt the search towards the most promising areas of the search space as it progresses. We use a state-of-the-art deep reinforcement learning algorithm, proximate policy optimization, to more efficiently search the large and continuous state space. We find that this approach significantly improves the performance of AST compared to a prior approach based on Monte Carlo tree search. We perform experiments using AST to find sNMAC events under various encounter configurations, varying parameters pertaining to dynamics and coordination. Our experiments show AST to be very effective at finding sNMAC scenarios. We summarize our findings, presenting high-level categories of discovered sNMACs and specific examples of encounters in each category.

aircraft collision avoidance↗

Integrated Process-Structure-Property Simulations for Additive Manufacturing Using the Open-Source Materialite Package

The microstructure and properties of additively manufactured (AM) metals are strongly dependent on process conditions. Therefore, process-structure-property (PSP) simulations are a useful tool for exploring process parameter space, studying process variations, and quantifying uncertainty in material properties. However, integrating process-structure and structure-property simulations often involves connecting multiple software packages. Each package may use unique data structures and require substantial domain knowledge. This presentation demonstrates PSP simulation capabilities of Materialite, an open-source package developed at NASA Langley Research Center. Materialite simplifies model linkages by using a common data structure and model interface, enabling straightforward simulation across a PSP model chain. Physics-based models, including kinetic Monte Carlo and crystal plasticity, are implemented within the package. The model interface is also intended to simplify implementation of new models and enable integration with external simulation tools. Example use cases include uncertainty quantification with PSP models and GPU-accelerated powder bed fusion AM process models.

additive manufacturing↗