Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “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 379 records · Page 21

Solving and Learning Soft Temporal Constraints: Experimental Scenario and Examples

Soft temporal constraint problems allow to describe in a natural way scenarios where events happen over time and preferences are associated to event distances and durations. However, sometimes such local preferences are difficult to set, and it may be easier instead to associate preferences to some complete solutions of the problem. To model everything in a uniform way via local preferences only, and also to take advantage of the existing constraint solvers which exploit only local preference use machine learning techniques which learn the local preferences from the global ones. In this paper we describe the existing framework for both solving and learning preferences in temporal constraint problems, the implemented modules, the experimental scenario, and preliminary results on some examples.

Rossi, F.↗

Collaborating Fuzzy Reinforcement Learning Agents

Earlier, we introduced GARIC-Q, a new method for doing incremental Dynamic Programming using a society of intelligent agents which are controlled at the top level by Fuzzy Relearning and at the local level, each agent learns and operates based on ANTARCTIC, a technique for fuzzy reinforcement learning. In this paper, we show that it is possible for these agents to compete in order to affect the selected control policy but at the same time, they can collaborate while investigating the state space. In this model, the evaluator or the critic learns by observing all the agents behaviors but the control policy changes only based on the behavior of the winning agent also known as the super agent.

Berenji, Hamid R.↗

Design and Control of Large Collections of Learning Agents

The intelligent control of multiple autonomous agents is an important yet difficult task. Previous methods used to address this problem have proved to be either too brittle, too hard to use, or not scalable to large systems. The 'Collective Intelligence' project at NASA/Ames provides an elegant, machine-learning approach to address these problems. This approach mathematically defines some essential properties that a reward system should have to promote coordinated behavior among reinforcement learners. This work has focused on creating additional key properties and algorithms within the mathematics of the Collective Intelligence framework. One of the additions will allow agents to learn more quickly, in a more coordinated manner. The other will let agents learn with less knowledge of their environment. These additions will allow the framework to be applied more easily, to a much larger domain of multi-agent problems.

Agogino, Adrian↗

MC-1 Engine Valves, Lessons Learned

Many lessons were learned during the development of the valves for the MC-1 engine. The purpose of this report is to focus on a variety of issues related to the engine valves and convey the lessons learned. This paper will not delve into detailed technical analysis of the components. None of the lessons learned are new or surprising, but simply reinforce the importance of addressing the details of the design early, at the component level. The Marshall Space Flight Center (MSFC), Huntsville, Alabama developed the MC-1 engine, a LOX / FW-1, 60,000 pound thrust engine. This engine was developed under the Low Cost Boost Technology office at MSFC and proved to be a very successful project for the MSFC Propulsion team and the various subcontractors working the development of the engine and its components.

Laszar, John↗

Web-Enhanced Instruction and Learning: Findings of a Short- and Long-Term Impact Study and Teacher Use of NASA Web Resources

Over a three-year period, researchers and educators from the Pennsylvania State University (PSU), University Park, Pennsylvania, and the NASA Dryden Flight Research Center (DFRC), Edwards, California, worked together to analyze, develop, implement and evaluate materials and tools that enable teachers to use NASA Web resources effectively for teaching science, mathematics, technology and geography. Two conference publications and one technical paper have already been published as part of this educational research series on Web-based instruction and learning. This technical paper, Web-Enhanced Instruction and Learning: Findings of a Short- and Long-Term Impact Study, is the culminating report in this educational research series and is based on the final report submitted to NASA. This report describes the broad spectrum of data gathered from teachers about their experiences using NASA Web resources in the classroom. It also describes participating teachers responses and feedback about the use of the NASA Web-Enhanced Learning Environment Strategies reflection tool on their teaching practices. The reflection tool was designed to help teachers merge the vast array of NASA resources with the best teaching methods, taking into consideration grade levels, subject areas and teaching preferences. The teachers described their attitudes toward technology and innovation in the classroom and their experiences and perceptions as they attempted to integrate Web resources into science, mathematics, technology and geography instruction.

McCarthy, Marianne C.↗

Propulsion Ground Testing with High Test Peroxide: Lessons Learned

Propulsion Ground Testing with High Test Peroxide (85 to 98% concentration) began at the NASA John C. Stennis Space Center in calendar year 1998, when the E3 Test Facility was modified to accomodate hydrogen peroxide (H2O2) in order to suport the research and development testing of the USAF Upper Stage Flight Experiment rocket engine. Since that time, efforts have continued to provide actual and planned test services to various customers, both U.S. Government and Commercial, in the ground test of many test articles, ranging from gas generators, to catalyst beds, to turbomachinery, to main injectors, to combustion chambers, to integrated rocket engines, to integrated stages. Along this path, and over the past 4 years, there has been both the rediscovery of previously learned lessons, through literature search, archive review, and personal interviews, as well as the learning of many new lessons as new areas are explored and new endeavors are tried. This paper will summarize those lessons learned in an effort to broaden the knowledge base as High Test Peroxide is considered more widely for use in rocket propulsion applications.

Bruce, Robert↗

Time-Extended Policies in Mult-Agent Reinforcement Learning

Reinforcement learning methods perform well in many domains where a single agent needs to take a sequence of actions to perform a task. These methods use sequences of single-time-step rewards to create a policy that tries to maximize a time-extended utility, which is a (possibly discounted) sum of these rewards. In this paper we build on our previous work showing how these methods can be extended to a multi-agent environment where each agent creates its own policy that works towards maximizing a time-extended global utility over all agents actions. We show improved methods for creating time-extended utilities for the agents that are both "aligned" with the global utility and "learnable." We then show how to crate single-time-step rewards while avoiding the pi fall of having rewards aligned with the global reward leading to utilities not aligned with the global utility. Finally, we apply these reward functions to the multi-agent Gridworld problem. We explicitly quantify a utility's learnability and alignment, and show that reinforcement learning agents using the prescribed reward functions successfully tradeoff learnability and alignment. As a result they outperform both global (e.g., team games ) and local (e.g., "perfectly learnable" ) reinforcement learning solutions by as much as an order of magnitude.

Tumer, Kagan↗

Spatial learning and memory is preserved in rats after early development in a microgravity environment

This study evaluated the cognitive mapping abilities of rats that spent part of their early development in a microgravity environment. Litters of male and female Sprague-Dawley rat pups were launched into space aboard the National Aeronautics and Space Administration space shuttle Columbia on postnatal day 8 or 14 and remained in space for 16 days. These animals were designated as FLT groups. Two age-matched control groups remained on Earth: those in standard vivarium housing (VIV) and those in housing identical to that aboard the shuttle (AGC). On return to Earth, animals were tested in three different tasks that measure spatial learning ability, the Morris water maze (MWM), and a modified version of the radial arm maze (RAM). Animals were also tested in an open field apparatus to measure general activity and exploratory activity. Performance and search strategies were evaluated in each of these tasks using an automated tracking system. Despite the dramatic differences in early experience, there were remarkably few differences between the FLT groups and their Earth-bound controls in these tasks. FLT animals learned the MWM and RAM as quickly as did controls. Evaluation of search patterns suggested subtle differences in patterns of exploration and in the strategies used to solve the tasks during the first few days of testing, but these differences normalized rapidly. Together, these data suggest that development in an environment without gravity has minimal long-term impact on spatial learning and memory abilities. Any differences due to development in microgravity are quickly reversed after return to earth normal gravity.

NASA Discipline Neuroscience↗

Detection in fixed and random noise in foveal and parafoveal vision explained by template learning

Foveal and parafoveal contrast detection thresholds for Gabor and checkerboard targets were measured in white noise by means of a two-interval forced-choice paradigm. Two white-noise conditions were used: fixed and twin. In the fixed noise condition a single noise sample was presented in both intervals of all the trials. In the twin noise condition the same noise sample was used in the two intervals of a trial, but a new sample was generated for each trial. Fixed noise conditions usually resulted in lower thresholds than twin noise. Template learning models are presented that attribute this advantage of fixed over twin noise either to fixed memory templates' reducing uncertainty by incorporation of the noise or to the introduction, by the learning process itself, of more variability in the twin noise condition. Quantitative predictions of the template learning process show that it contributes to the accelerating nonlinear increase in performance with signal amplitude at low signal-to-noise ratios.

NASA Center ARC↗

Behavioral analysis of signals that guide learned changes in the amplitude and dynamics of the vestibulo-ocular reflex

We characterized the dependence of motor learning in the monkey vestibulo-ocular reflex (VOR) on the duration, frequency, and relative timing of the visual and vestibular stimuli used to induce learning. The amplitude of the VOR was decreased or increased through training with paired head and visual stimulus motion in the same or opposite directions, respectively. For training stimuli that consisted of simultaneous pulses of head and target velocity 80-1000 msec in duration, brief stimuli caused small changes in the amplitude of the VOR, whereas long stimuli caused larger changes in amplitude as well as changes in the dynamics of the reflex. When the relative timing of the visual and vestibular stimuli was varied, brief image motion paired with the beginning of a longer vestibular stimulus caused changes in the amplitude of the reflex alone, but the same image motion paired with a later time in the vestibular stimulus caused changes in the dynamics as well as the amplitude of the VOR. For training stimuli that consisted of sinusoidal head and visual stimulus motion, low-frequency training stimuli induced frequency-selective changes in the VOR, as reported previously, whereas high-frequency training stimuli induced changes in the amplitude of the VOR that were more similar across test frequency. The results suggest that there are at least two distinguishable components of motor learning in the VOR. One component is induced by short-duration or high-frequency stimuli and involves changes in only the amplitude of the reflex. A second component is induced by long-duration or low-frequency stimuli and involves changes in the amplitude and dynamics of the VOR.

NASA Discipline Neuroscience↗

Atmospheric/Space Environment Support Lessons Learned Regarding Aerospace Vehicle Design and Operations

In modern government and aerospace industry institutions the necessity of controlling current year costs often leads to high mobility in the technical workforce, "one-deep" technical capabilities, and minimal mentoring for young engineers. Thus, formal recording, use, and teaching of lessons learned are especially important in the maintenance and improvement of current knowledge and development of new technologies, regardless of the discipline area. Within the NASA Technical Standards Program Website http://standards.nasa.gov there is a menu item entitled "Lessons Learned/Best Practices". It contains links to a large number of engineering and technical disciplines related data sets that contain a wealth of lessons learned information based on past experiences. This paper has provided a small sample of lessons learned relative to the atmospheric and space environment. There are many more whose subsequent applications have improved our knowledge of the atmosphere and space environment, and the application of this knowledge to the engineering and operations for a variety of aerospace programs.

Vaughan, William W.↗

Learning Science Through Visualization

In the context of an introductory physical science course for non-science majors, I have been trying to understand how scientific visualizations of natural phenomena can constructively impact student learning. I have also necessarily been concerned with the instructional and assessment approaches that need to be considered when focusing on learning science through visually rich information sources. The overall project can be broken down into three distinct segments : (i) comparing students' abilities to demonstrate proportional reasoning competency on visual and verbal tasks (ii) decoding and deconstructing visualizations of an object falling under gravity (iii) the role of directed instruction to elicit alternate, valid scientific visualizations of the structure of the solar system. Evidence of student learning was collected in multiple forms for this project - quantitative analysis of student performance on written, graded assessments (tests and quizzes); qualitative analysis of videos of student 'think aloud' sessions. The results indicate that there are significant barriers for non-science majors to succeed in mastering the content of science courses, but with informed approaches to instruction and assessment, these barriers can be overcome.

Chaudhury, S. Raj↗

Importance Of Quality Control in Reducing System Risk, a Lesson Learned From The Shuttle and a Recommendation for Future Launch Vehicles

This paper presents lessons learned from the Space Shuttle return to flight experience and the importance of these lessons learned in the development of new the NASA Crew Launch Vehicle (CLV). Specifically, the paper discusses the relationship between process control and system risk, and the importance of process control in improving space vehicle flight safety. It uses the External Tank (ET) Thermal Protection System (TPS) experience and lessons learned from the redesign and process enhancement activities performed in preparation for Return to Flight after the Columbia accident. The paper also, discusses in some details, the Probabilistic engineering physics based risk assessment performed by the Shuttle program to evaluate the impact of TPS failure on system risk and the application of the methodology to the CLV.

Safie, Fayssal M.↗

NASA Crew and Cargo Launch Vehicle Development Approach Builds on Lessons Learned from Past and Present Missions

A viewgraph presentation of NASA crew and cargo launch vehicle development building upon lessons learned from past and present missions is shown. The topics include: 1) U.S. Vision for Space Exploration; 2) NASA's Exploration Road Map; 3) The Moon-The First Step to Mars and Beyond; 4) Building on a Foundation of Proven Technologies; 5) Constellation Launch Vehicle Elements; 6) 1.5 Launch Earth Orbit/Lunar Orbit Rendezvous; 7) The Journey Continues; 8) Design Philosophy for Mission Success; 9) Lessons Learned: Implementation Tenets; and 10) Lessons Learned: Early Integration with the Operators.

Dumbacher, Dan↗

Habitability and Human Factors: Lessons Learned in Long Duration Space Flight

This study documents the investigation of qualitative habitability and human factors feedback provided by scientists, engineers, and crewmembers on lessons learned from the ISS Program. A thorough review and understanding of this data is critical in charting NASA's future path in space exploration. NASA has been involved in ensuring that the needs of crewmembers to live and work safely and effectively in space have been met throughout the ISS Program. Human factors and habitability data has been collected from every U.S. crewmember that has resided on the ISS. The knowledge gained from both the developers and inhabitants of the ISS have provided a significant resource of information for NASA and will be used in future space exploration. The recurring issues have been tracked and documented; the top 5 most critical issues have been identified from this data. The top 5 identified problems were: excessive onsrbit stowage; environment; communication; procedures; and inadequate design of systems and equipment. Lessons learned from these issues will be used to aid in future improvements and developments to the space program. Full analysis of the habitability and human factors data has led to the following recommendations. It is critical for human factors to be involved early in the design of space vehicles and hardware. Human factors requirements need to be readdressed and redefined given the knowledge gained during previous ISS and long-duration space flight programs. These requirements must be integrated into vehicle and hardware technical documentation and consistently enforced. Lastly, space vehicles and hardware must be designed with primary focus on the user/operator to successfully complete missions and maintain a safe working environment. Implementation of these lessons learned will significantly improve NASA's likelihood of success in future space endeavors.

Baggerman, Susan D.↗

Lessons Learned and Flight Results from the F15 Intelligent Flight Control System Project

A viewgraph presentation on the lessons learned and flight results from the F15 Intelligent Flight Control System (IFCS) project is shown. The topics include: 1) F-15 IFCS Project Goals; 2) Motivation; 3) IFCS Approach; 4) NASA F-15 #837 Aircraft Description; 5) Flight Envelope; 6) Limited Authority System; 7) NN Floating Limiter; 8) Flight Experiment; 9) Adaptation Goals; 10) Handling Qualities Performance Metric; 11) Project Phases; 12) Indirect Adaptive Control Architecture; 13) Indirect Adaptive Experience and Lessons Learned; 14) Gen II Direct Adaptive Control Architecture; 15) Current Status; 16) Effect of Canard Multiplier; 17) Simulated Canard Failure Stab Open Loop; 18) Canard Multiplier Effect Closed Loop Freq. Resp.; 19) Simulated Canard Failure Stab Open Loop with Adaptation; 20) Canard Multiplier Effect Closed Loop with Adaptation; 21) Gen 2 NN Wts from Simulation; 22) Direct Adaptive Experience and Lessons Learned; and 23) Conclusions

Bosworth, John↗

Lessons Learned from Seven Space Shuttle Missions

Much can be learned from well-written descriptions of the technical and organizational factors that lead to an accident. Subsequent analysis by third parties of investigation reports and associated evidence collected during the investigations can lead to additional insight. Much can also be learned from documented close calls that do not result in loss of life or a spacecraft, such as the Mars Exploration Rover Spirit software anomaly, the SOHO mission interruption, and the NEAR burn anomaly. Seven space shuttle incidents fall into the latter category: Rendezvous Target Failure On STS-41B; Rendezvous Radar Anomaly and Trajectory Dispersion-STS-32 ;Rendezvous Lambert Targeting Anomaly-STS-49; Rendezvous Lambert Targeting Anomaly-STS-51; Zero Doppler Steering Maneuver Anomaly-STS-59; Excessive Propellant Consumption During Rendezvous-STS-69; Global Positioning System Receiver and Associated Shuttle Flight Software Anomalies-STS-91 Procedural work-arounds or software changes prevented them from threatening mission success. Extensive investigations, which included the independent recreation of the anomalies by multiple Shuttle Program organizations, were the key to determining the cause, accurately assessing risk, and identifying software and software process improvements. Lessons learned from these incidents not only validated long-standing operational best practices, but serve to promote discussion and mentoring among Program personnel and are applicable to future space flight programs.

Goodman, John↗

Learning Grasp Strategies Composed of Contact Relative Motions

Of central importance to grasp synthesis algorithms are the assumptions made about the object to be grasped and the sensory information that is available. Many approaches avoid the issue of sensing entirely by assuming that complete information is available. In contrast, this paper proposes an approach to grasp synthesis expressed in terms of units of control that simultaneously change the contact configuration and sense information about the object and the relative manipulator-object pose. These units of control, known as contact relative motions (CRMs), allow the grasp synthesis problem to be recast as an optimal control problem where the goal is to find a strategy for executing CRMs that leads to a grasp in the shortest number of steps. An experiment is described that uses Robonaut, the NASA-JSC space humanoid, to show that CRMs are a viable means of synthesizing grasps. However, because of the limited amount of information that a single CRM can sense, the optimal control problem may be partially observable. This paper proposes expressing the problem as a k-order Markov Decision Process (MDP) and solving it using Reinforcement Learning. This approach is tested in a simulation of a two-contact manipulator that learns to grasp an object. Grasp strategies learned in simulation are tested on the physical Robonaut platform and found to lead to grasp configurations consistently.

Platt, Robert, Jr.↗