Search

UNKNOWN

UNKNOWN

UNKNOWN

UNKNOWN

UNKNOWN

UNKNOWN

Recent commercial developments in multicore processors (e.g. Tilera, Clearspeed, HyperX) have provided an option for high performance embedded computing that rivals the performance attainable with FPGA-based reconfigurable computing architectures. Furthermore, these processors offer more straightforward and streamlined application development by allowing the use of conventional programming languages and software tools in lieu of hardware design languages such as VHDL and Verilog. With these advantages, multicore processors can significantly enhance the capabilities of future robotic space missions. This paper will discuss these benefits, along with onboard processing applications where multicore processing can offer advantages over existing or competing approaches. This paper will also discuss the key artchitecural features of current commercial multicore processors. In comparison to the current art, the features and advancements necessary for spaceflight multicore processors will be identified. These include power reduction, radiation hardening, inherent fault tolerance, and support for common spacecraft bus interfaces. Lastly, this paper will explore how multicore processors might evolve with advances in electronics technology and how avionics architectures might evolve once multicore processors are inserted into NASA robotic spacecraft.

A viewgraph presentation about information technology and its role in space missions is shown. The topics include: 1) Where is the IT on Space Missions? 2) Winners of the NASA Software of the Year Award; 3) Space Networking Roadmap; and 4) 10 (7) -Year Vision for IT in Space.

Space exploration missions are undergoing a significant transformation as are the expectations of their scientific investigators and the public who participate in these great voyages of exploration.

This paper discusses the following topics: (1) Autonomy for Future Missions- Mars Outposts, Titan Aerobot, and Europa Cryobot / Hydrobot; (2) Emergence of Autonomy- Remote Agent Architecture, Closing Loops Onboard, and New Millennium Flight Experiment; and (3) Software Engineering Challenges- Influence of Remote Agent, Scalable Autonomy, Autonomy Software Validation, Analytic Verification Technology, and Autonomy and Software Software Engineering.

The development of autonomy capabilities is the key to three vastly important strategic technical challenges facing NASA: the reduction of mission costs, the continuing return of quality science products through limited communications bandwidth, and the launching of a new era of solar system exploration characterized by sustained presence and in-depth scientific studies, including the search for life. Autonomy will benefit future NASA missions by migrating routine, traditionally ground-based functions to the spacecraft, by directly supporting the decoupling of spacecraft from the ground through new operations concepts, by enabling direct links between scientists and the space platforms carrying their instruments of investigation, and by the closing, of planning and control loops onboard, enabling space platforms to directly address uncertainty in the real-time mission context. The talk will survey ongoing, autonomy technology development projects at NASA, many of which have been or will soon be the subject of flight technology experiments, or are already targeted for mission use. The talk will also survey the exciting suite of future NASA space exploration missions, and make the case for the central role of autonomy in achieving the goals of these bold, unprecedented missions: cooperating rovers on the surface of Mars, the search for Earth-like planets around nearby stars, asteroid and comet landers, aerobots in planetary atmospheres, and a series of missions to intriguing Europa, perhaps culminating in a submersible to investigate its putative ocean. Finally, the talk will conclude with some farther-reaching speculations on how to create properties such as long-term survivability and evolvability in future space systems, such that they will be well equipped to extend humanity exploratory presence into the interstellar realm.

The challenge of space flight in NASA's future is to enable more frequent and more intensive space exploration missions at lower cost.

Any attempt to introduce automation into the monitoring of complex physical systems must start from a robust anomaly detection capability. This task is far from straightforward, for a single definition of what constitutes an anomaly is difficult to come by. In addition, to make the monitoring process efficient, and to avoid the potential for information overload on human operators, attention focusing must also be addressed. When an anomaly occurs, more often than not several sensors are affected, and the partially redundant information they provide can be confusing, particularly in a crisis situation where a response is needed quickly. The focus of this paper is a new technique for attention focusing. The technique involves reasoning about the distance between two frequency distributions, and is used to detect both anomalous system parameters and 'broken' causal dependencies. These two forms of information together isolate the locus of anomalous behavior in the system being monitored.

In real-time monitoring situations, more information is not necessarily better. When faced with complex emergency situations, operators can experience information overload and a compromising of their ability to react quickly and correctly. We describe an approach to focusing operator attention in real-time systems monitoring based on a set of empirical and model-based measures for determining the relative importance of sensor data.

Scientific data preparation is the process of extracting usable scientific data from raw instrument data. This task involves noise detection (and subsequent noise classification and flagging or removal), extracting data from compressed forms, and construction of derivative or aggregate data (e.g. spectral densities or running averages). A software system called PIPE provides intelligent assistance to users developing scientific data preparation plans using a programming language called Master Plumber. PIPE provides this assistance capability by using a process description to create a dependency model of the scientific data preparation plan. This dependency model can then be used to verify syntactic and semantic constraints on processing steps to perform limited plan validation. PIPE also provides capabilities for using this model to assist in debugging faulty data preparation plans. In this case, the process model is used to focus the developer's attention upon those processing steps and data elements that were used in computing the faulty output values. Finally, the dependency model of a plan can be used to perform plan optimization and run time estimation. These capabilities allow scientists to spend less time developing data preparation procedures and more time on scientific analysis tasks.

Within this decade, the growth in complexity of exploratory data analysis and the sheer volume of space data require new and innovative approaches to support science investigators in achieving their research objectives. To date, there have been numerous efforts addressing the individual issues involved in inter-disciplinary, multi-instrument investigations. However, while successful in small scale, these efforts have not proven to be open and scalable. This proposal addresses four areas of significant need: scientific visualization and analysis; science data management; interactions in a distributed, heterogeneous environment; and knowledge-based assistance for these functions. The fundamental innovation embedded with this proposal is the integration of three automation technologies, namely, knowledge-based expert systems, science visualization and science data management. This integration is based on concept called the DataHub. With the DataHub concept, NASA will be able to apply a more complete solution to all nodes of a distributed system. Both computation nodes and interactive nodes will be able to effectively and efficiently use the data services (address, retrieval, update, etc.) with a distributed, interdisciplinary information system in a uniform and standard way. This will allow the science investigators to concentrate on their scientific endeavors, rather than to involve themselves in the intricate technical details of the systems and tools required to accomplish their work. Thus, science investigators need not be programmers. The emphasis will be on the definition and prototyping of system elements with sufficient detail to enable data analysis and interpretation leading to publishable scientific results. In addition, the proposed work includes all the required end-to-end components and interfaces to demonstrate the completed concept.

We describe the application of machine classification techniques to the development of an automated tool for the reduction of a large scientific data set. The 2nd Palomer Observatory Sky Survey is nearly completed. This survey provides comprehensive coverage of the northern celestial hemisphere in the form of photographic plates. The plates are being transformed into digitized images whose quality will probably not be surpassed in the next ten to twenty years. The images are expected to contain on the order of 10(exp 7) galaxies and 10(exp 8) stars. Astronomers wish to determine which of these sky objects belong to various classes of galaxies and stars. The size of this data set precludes manual analysis. Our approach is to develop a software system which integrates the functions of independently developed techniques for image processing and data classification. Digitized sky images are passed through image processing routines to identify sky objects and to extract a set of features for each object. These routines are used to help select a useful set of attributes for classifying sky objects. Then GID3* and O-BTree, two inductive learning techniques, learn classification decision trees from examples. These classifiers will be used to process the rest of the data. This paper gives an overview of the machine learning techniques used, describes the details of our specific application, and reports the initial encouraging results. The results indicate that our approach is well-suited to the problem. The primary benefits of the approach are increased data reduction throughput and consistency of classification. The classification rules which are the product of the inductive learning techniques will form an object, examinable basis for classifying sky objects. A final, not to be underestimated benefit is that astronomers will be freed from the tedium of an intensely visual task to pursue more challenging analysis and interpretation problems based on automatically cataloged data.

Proposed method of predictive monitoring that prevents large number of sensor data from overwhelming human operator or electronic monitor is subject of continuing research. Predictive-monitoring system concentrates limited monitoring resources on few of many sensor outputs that are most important. Casual importance serves as basis for selection of sensors to verify expected behavior of system. Alarms raised where discrepancies between predicted and actual values are significant.

Within this decade, the growth in complexity of exploratory data analysis and the sheer volume of space data require new and innovative approaches to support science investigators in achieving their research objectives. To date, there have been numerous efforts addressing the individual issues involved in inter-disciplinary, multi-instrument investigations. However, while successful in small scale, these efforts have not proven to be open and scaleable. This proposal addresses four areas of significant need: scientific visualization and analysis; science data management; interactions in a distributed, heterogeneous environment; and knowledge-based assistance for these functions. The fundamental innovation embedded within this proposal is the integration of three automation technologies, namely, knowledge-based expert systems, science visualization and science data management. This integration is based on the concept called the Data Hub. With the Data Hub concept, NASA will be able to apply a more complete solution to all nodes of a distributed system. Both computation nodes and interactive nodes will be able to effectively and efficiently use the data services (access, retrieval, update, etc.) with a distributed, interdisciplinary information system in a uniform and standard way. This will allow the science investigators to concentrate on their scientific endeavors, rather than to involve themselves in the intricate technical details of the systems and tools required to accomplish their work. Thus, science investigators need not be programmers. The emphasis will be on the definition and prototyping of system elements with sufficient detail to enable data analysis and interpretation leading to publishable scientific results. In addition, the proposed work includes all the required end-to-end components and interfaces to demonstrate the completed concept.

On Space Station Freedom (SSF), design iterations have made clear the need to keep the sensor complement small. Along with the unprecendented duration of the mission, it is imperative that decisions regarding placement of sensors be carefully examined and justified during the design phase. In the ECLSS Predictive Monitoring task, we are developing AI-based software to enable design engineers to evaluate alternate sensor configurations. Based on techniques from model-based reasoning and information theory, the software tool makes explicit the quantitative tradeoffs among competing sensor placements, and helps designers explore and justify placement decisions. This work is being applied to the Environmental Control and Life Support System (ECLSS) testbed at MSFC to assist design personnel in placing sensors for test purposes to evaluate baseline configurations and ultimately to select advanced life support system technologies for evolutionary SSF.

Viewgraphs on DataHub knowledge based assistance for science visualization and analysis using large distributed databases. Topics covered include: DataHub functional architecture; data representation; logical access methods; preliminary software architecture; LinkWinds; data knowledge issues; expert systems; and data management.

A Spacecraft Health Automated Reasoning Prototype (SHARP) portability study is presented. Some specific progress is described on the portability studies, plans for technology transfer, and potential applications of SHARP and related artificial intelligence technology to telescience operations. The application of SHARP to Voyager telecommunications was a proof-of-capability demonstration of artificial intelligence as applied to the problem of real time monitoring functions in planetary mission operations. An overview of the design and functional description of the SHARP system is also presented as it was applied to Voyager.

An approach to selective processing in monitoring is described. The approach is designed to provide informative feedback on whether a system is performing nominally in the current operating context without exceeding available resources for prediction and interpretation. At the center of the approach to selective sensor interpretation and simulation is a causal model of the system being monitored. The architecture of the monitoring system, called PREMON, is described as well as how causal models of physical systems are represented and simulated.

The quick and reliable detection of anomalous behavior in aerospace systems is addressed. Ways to adjust nominal sensor value expectations dynamically, taking into account the changing operating context of the system, are considered. How to utilize sensors selectively, determining which subset of the available sensors to use at any given time to verify nominal operation efficiently without processing a prohibitive amount of data, is examined.

Two issues are addressed which arise in the task of detecting anomalous behavior in complex systems with numerous sensor channels: how to adjust alarm thresholds dynamically, within the changing operating context of the system, and how to utilize sensors selectively, so that nominal operation can be verified reliably without processing a prohibitive amount of sensor data. The approach involves simulation of a causal model of the system, which provides information on expected sensor values, and on dependencies between predicted events, useful in assessing the relative importance of events so that sensor resources can be allocated effectively. The potential applicability of this work to the execution monitoring of robot task plans is briefly discussed.

Traditional approaches to monitoring are proving inadequate in the face of two important issues: the dynamic adjustment of expectations about sensor values when the behavior of the device is too complex to enumerate beforehand, and the selective but effective interpretation of sensor readings when the number of sensors becomes overwhelming. This system addresses these issues by building an explicit model of a device and applying common-sense theories of physics to model causality in the device. The resulting causal simulation of the device supports planning decisions about how to efficiently yet reliably utilize a limited number of sensors to verify correct operation of the device.