SEARCH · Search NASA
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.
We describe a progression from pilot studies to development and use of domain-specific verification and validation (V&V) automation. Our domain is the testing of an AI planning system that forms a key component of an autonomous spacecraft.
Research under this grant successfully developed and tested the Interruptable Control System (ICE). The best description of ICE is in the M.S. thesis of Jim Vezina, who was the primary designer and implementer of ICE. Vezina's thesis is included as the bulk of this report. To test the general usefulness and maintainability of the ICE System, the examples of the ICE system in action given in Vezina's thesis were extended. The extensions consisted of two components: debugging and rewriting the simulation of passive and active agents, both controlled and uncontrolled, and benchmarking more extensively the HES expert system concerned with monitoring sensors. The expert system was chosen because it demonstrates the ability of CLIPS.
Interaction in heterogeneous mission operations teams is not well matched to classical models of coordination with autonomous systems. We describe methods of loose coordination and information management in mission operations. We describe an information agent and information management tool suite for managing information from many sources, including autonomous agents. We present an integrated model of levels of complexity of agent and human behavior, which shows types of information processing and points of potential error in agent activities. We discuss the types of information needed for diagnosing problems and planning interactions with an autonomous system. We discuss types of coordination for which designs are needed for autonomous system functions.
To achieve the goal of building an autonomous agent, the usually disjoint capabilities of planning, execution, and learning must be used together. An architecture, called MAX, within which cognitive capabilities can be purposefully and intelligently integrated is described. The architecture supports the codification of capabilities as explicit knowledge that can be reasoned about. In addition, specific problem solving, learning, and integration knowledge is developed.
We discuss the opportunities for autonomous sys- tems to perform reflection on their planners by adapting the models used to build plans. We first describe model-based planning systems, a form of automated planning system driven by declarative models of the planning domain. These models include descriptions of the conditions and effects of actions on the state of the world. When planning the activities of cyber-physical systems, the command and data representation of the system must be formally abstracted to the actions and states described in the planning system model. When the execution of a plan either fails or produces unexpected outcomes, the execution trace can be abstracted and compared to the predicted state according to the planning model, producing a list of discrepancies; these discrepancies can then be used to fix the model. This provides part of a reflection capability, namely, a set of well-formed problems with the domain model, the abstractions, or both. The challenge lies in the rest of the reflection capability, namely, a set of techniques for changing the models or the abstractions. We discuss these challenges and describe some of the options for addressing them
This study, now in its third year, has had the overall objective and challenge of determining the needed hooks and scars in the initial Space Station Freedom (SSF) system to assure that on-orbit assembly and refurbishment of lunar and Mars spacecraft can be accomplished with the maximum use of automation. In this study automation is all encompassing and includes physical tasks such as parts mating, tool operation, and human visual inspection, as well as non-physical tasks such as monitoring and diagnosis, planning and scheduling, and autonomous visual inspection. Potential tasks for automation include both extravehicular activity (EVA) and intravehicular activity (IVA) events. A number of specific techniques and tools have been developed to determine the ideal tasks to be automated, and the resulting timelines, changes in labor requirements and resources required. The Mars/Phobos exploratory mission developed in FY89, and the Lunar Assembly/Refurbishment mission developed in FY90 and depicted in the 90 Day Study as Option 5, have been analyzed in detailed in recent years. The complete methodology and results are presented in FY89 and FY90 final reports.
This paper describes the plans for an experiment undertaken by the TOPEX/POSEIDON (T/P) mission to demonstrate autonomous on-board manuever design and implementation.
A general definition of Mission Planning is given. The definition covers the full scope of an end-to-end mission planning system. Noting the mission-specific nature of most mission planning systems, a classification of autonomous spacecraft missions is made into Observatory, Survey, multi-instrument science, and Telecommunications missions. The mission planning approach for one mission in each category is examined critically. The following missions were chosen: ISO (Infrared Space Observatory); ERS-1 (European Remote Sensing Satellite); and Eureca (European Retrievable Carrier). The paper gives a summary of lessons learned from these missions suggesting improvements in areas such as requirements analysis, testing, user interfacing, rules, and constraints handling. The paper will also examine commonalities in functions, which could constitute a basis for identification of generic mission planning support tools.
NASA's Strategic Plan for the Aerospace Technology Enterprise includes ambitious objectives focused on affordable air travel, reduced emissions, and expanded aviation-system capacity. NASA Dryden Flight Research Center, in cooperation with NASA Ames Research Center, the Boeing Company, and the University of California, Los Angeles, has embarked on an autonomous-formation-flight project that promises to make significant strides towards these goals. For millions of years, birds have taken advantage of the aerodynamic benefit of flying in formation. The traditional "V" formation flown by many species of birds (including gulls, pelicans, and geese) enables each of the trailing birds to fly in the upwash flow field that exists just outboard of the bird immediately ahead in the formation. The result for each trailing bird is a decrease in induced drag and thus a reduction in the energy needed to maintain a given speed. Hence, for migratory birds, formation flight extends the range of the system of birds over the range of birds flying solo. The Autonomous Formation Flight (AFF) Project is seeking to extend this symbiotic relationship to aircraft.
An introduction is given to the Value-driven methodology, which has been successfully applied to solve a variety of difficult decision, control, and optimization problems. Many real-world decision processes (e.g., those encountered in scheduling, allocation, and command and control) involve a hierarchy of complex planning considerations. For such problems it is virtually impossible to define a fixed set of rules that will operate satisfactorily over the full range of probable contingencies. Decision Science Applications' value-driven methodology offers a systematic way of automating the intuitive, common-sense approach used by human planners. The inherent responsiveness of value-driven systems to user-controlled priorities makes them particularly suitable for semi-automated applications in which the user must remain in command of the systems operation. Three examples of the practical application of the approach in the automation of hierarchical decision processes are discussed: the TAC Brawler air-to-air combat simulation is a four-level computerized hierarchy; the autonomous underwater vehicle mission planning system is a three-level control system; and the Space Station Freedom electrical power control and scheduling system is designed as a two-level hierarchy. The methodology is compared with rule-based systems and with other more widely-known optimization techniques.
The Autonomous Sciencecraft Experiment (ASE), currently flying onboard the Earth Observing-1 (EO-I) spacecraft, integrates several autonomy software technologies enabling autonomous science analysis and mission planning. The experiment demonstrates the potential for future space missions to use onboard decision-making to respond autonomously to capture short-lived science phenomena. The AAAI software demonstration will consist of two sections: a real-time display of an ASE-commanded ground contact from the EO-1 spacecraft, and a simulation of the full ASE autonomous science-response scenario.
Because of communication limits, planetary rovers must operate autonomously during consequent durations. The ability to plan under uncertainty is one of the main components of autonomy. Previous approaches to planning under uncertainty in NASA applications are not able to address the challenges of future missions, because of several apparent limits. On another side, decision theory provides a solid principle framework for reasoning about uncertainty and rewards. Unfortunately, there are several obstacles to a direct application of decision-theoretic techniques to the rover domain. This paper focuses on the issues of structure and concurrency, and continuous state variables. We describes two techniques currently under development that address specifically these issues and allow scaling-up decision theoretic solution techniques to planetary rover planning problems involving a small number of goals.
NASA and military applications for high power spacecraft are presently being planned. A perspective of autonomous power management for low earth orbit large space platforms is presented. A multichannel 250 kWe utility-type power subsystem is used as a baseline system. The need for automation is reviewed, based on power subsystem complexity, survivability requirements, and cost benefits. On-board versus ground management is discussed with respect to these needs. In addition, the utilization of autonomous power management to enable technology readiness of large complex power subsystems is described. Recommendations are made for further technology thrusts.
The National Aeronautics and Space Administration (NASA) Space Station Program marks the begining of a new era in space utilization and habitation. Extensive use of remote manipulation and robotics to reduce astronaut extra-vehicular activity is expected. Emphasis on teleoperator technology in early Space Station phases, followed by growth of autonomous robotics capabilities, is planned. A new telerobot concept has been developed at Oak Ridge National Laboratory (ORNL) under NASA Langley Research Center sponsorship, to address the technical needs of both teleoperations and telerobotics for these future NASA programs. The concept is based on traction drives, redundant kinematics, modular construction, and a state-of-the-art distributed, hierarchical control system.
A hierarchical functional autonomous controller architecture is introduced. In particular, the architecture for the control of future space vehicles is described in detail; it is designed to ensure the autonomous operation of the control system and it allows interaction with the pilot and crew/ground station, and the systems on board the autonomous vehicle. The fundamental issues in autonomous control system modeling and analysis are discussed. It is proposed to utilize a hybrid approach to modeling and analysis of autonomous systems. This will incorporate conventional control methods based on differential equations and techniques for the analysis of systems described with a symbolic formalism. In this way, the theory of conventional control can be fully utilized. It is stressed that autonomy is the design requirement and intelligent control methods appear at present, to offer some of the necessary tools to achieve autonomy. A conventional approach may evolve and replace some or all of the `intelligent' functions. It is shown that in addition to conventional controllers, the autonomous control system incorporates planning, learning, and FDI (fault detection and identification).
An overview of MAX, a software framework for manager's associate systems, is presented. MAX is used to develop and execute a problem-solving strategy for the task planning of semi-autonomous agents with the assistance of human performance. This paper describes the use of MAX in the supervisory management of robotic vehicles as they explore a planetary surface.
This paper describes an integrated demonstration of ground-based contingent planning, robust execution and autonomous instrument placement for the efficient exploration of a site by a prototype Mars rover.
A viewgraph presentation describing an analytical thermal model used to enable research on autonomous soaring for a small UAV aircraft is given. The topics include: 1) Purpose; 2) Approach; 3) SURFRAD Data; 4) Convective Layer Thickness; 5) Surface Heat Budget; 6) Surface Virtual Potential Temperature Flux; 7) Convective Scaling Velocity; 8) Other Calculations; 9) Yearly trends; 10) Scale Factors; 11) Scale Factor Test Matrix; 12) Statistical Model; 13) Updraft Strength Calculation; 14) Updraft Diameter; 15) Updraft Shape; 16) Smoothed Updraft Shape; 17) Updraft Spacing; 18) Environment Sink; 19) Updraft Lifespan; 20) Autonomous Soaring Research; 21) Planned Flight Test; and 22) Mixing Ratio.