Thesis: S.M., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2017. / Cataloged from PDF version of thesis. / Includes bibliographical references (page 60). / Required Navigation Performance (RNP) instrument procedures guarantee high levels of navigation precision through highly accurate navigation sources (e.g. GPS) and real-time monitoring of position estimation accuracy. In recent years, the Federal Aviation Administration (FAA) has developed and published public RNP approach procedures at airports across the country. These RNP procedures offer unique capabilities such as curved segments (radius-to-fix, or RF legs), narrow containment areas, and constant descent profiles that are not seen combined in other categories of instrument approaches. Because of these capabilities, RNP approaches are regarded as highly flexible procedures that can be designed to meet specific stakeholder requirements (e.g. lower minimums in mountainous areas, minimizing fuel bum during approach, avoiding flight over populated areas for noise abatement, etc.) at the airport level. Among the various proposed benefits of RNP approaches, this study analyzed potential safety benefits related to improvements in approach stability. In total, 11,062 individual approaches at four airports were analyzed using radar (ASDE-X) data, of which 364 (3.29%) were identified as RNP procedures. Of all approaches analyzed, two non-RNP cases were identified as unsafe, while there were no unsafe RNP cases. However, due to the relatively low number of RNP approaches observed, no statistically significant evidence of improved stability on RNP approaches was found. Given the low utilization of RNP approach procedures found from radar data, further work was done to identify barriers to operational use of these procedures and to investigate strategies to accelerate the adoption of RNP across the National Airspace System (NAS). Potential factors driving the low utilization of RNP procedures were found to be the low levels of equipage and operational approval among air carriers, and difficulties in air traffic management stemming from mixed equipage operations. / by Sandro Salgueiro Rodrigues Filho. / S.M.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2000. / Also available online at the MIT Theses Online homepage <http://thesis.mit.edu>. / Includes bibliographical references (p. 109-112). / Artificial gravity is made through the centripetal force from a rotating chair or short-radius centrifuge. It is a very promising countermeasure, as it alone should remove all the adverse effects of microgravity. In order to effectively use artificial gravity as a long-duration space flight countermeasure, the effects of artificial gravity on the human body must be investigated. If artificial gravity is created by use of a short-radius centrifuge, the high angular velocity required, about 23 rpm, causes unexpected and illusory body motions when making head turns. My work in artificial gravity consisted of two parts, a study that investigated the vestibular response to head movements during centrifugation and regulator feedback control of the centrifuge. This experiment studied the perceived illusory body sensations and heart rate changes induced by head movements in both the yaw and pitch planes while supine during centrifugation. Yaw right, yaw left, and pitch head movements yielded successively significantly higher heart rate than baseline. Results show that 68% of subjects in the yaw plane and 48% of subjects in the pitch plane experienced illusory body tilt as predicted by a model of the vestibular system while 13% in yaw and 40% in pitch experienced body tilt in the opposite direction from the predicted model. Pitch head movements yielded significantly higher magnitude and duration of illusory tilt. These side effects are serious and will need to be controlled if short-radius centrifugation is to be a successful countermeasure. Regulator feedback control has been implemented on the centrifuge with both an optical encoder and an accelerometer. Tachometer development, automatic control, and classical PID control theory was used to develop the gain and integrator time constants, which lead to K=1.5 and Ti=1 sec. This results in an improved steady state error by 99.8% and a more accurate response of the centrifuge by 5.7% for the accelerometer and 52% for the encoder feedback system from the open loop system. / by Carol C. Cheung. / S.M.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2007. / Includes bibliographical references (p. 131-133). / Physical information is transferred between technical systems through wires, beams, and other physical attributes, while more intangible information is typically transferred between communities of people through artifacts such as documents, e-mails, etc. This research attempts to characterize these communication interfaces better by analyzing the use of artifacts at these interfaces by means of a boundary object attribute model. Boundary objects, the metric of analysis of this thesis, are artifacts used to bridge information and knowledge gaps between different communities of practice. The US Army's Future Combat System (FCS) was chosen as a case study primarily because of its complex programmatic characteristics. The information gathered in the FCS case study was combined with knowledge from previous boundary object literature to generate an attributes model. Once developed, the boundary object attributes model was validated on the US Air Force Transformational Communications Satellite System (TSAT) program focusing specifically on the TSAT Mission Operations System (TMOS) segment of the program. Data were collected on the frequency and type of resources used to understand information and the dependencies that individuals have with each other for documented information. Furthermore, five communication artifacts were critiqued for their effectiveness as boundary objects. Statistical tests were conducted to highlight trends in resource dependencies and attributes common in effective boundary objects. An implication of this research is that the most important attributes for a boundary object are inclusivity, traceability, and synchronization. This research also found that people generally tend to rely much more on other people for information than artifacts. This introduces problems of exhausting valuable human resources and creating unnecessary bottlenecks. / (cont.) A second implication of this research is that spending the extra time and effort to design artifacts with high inclusivity and freshness will add significant value to the overall system. In addition, a third implication of this research is that having the right boundary objects alone is not enough for effective collaborative interfaces. A fourth implication of this research is that designing a boundary object whose form follows its function is critical for its effectiveness. These suggestions can provide relief to a program highly taxing to its human resources and reduce transaction costs of the overall system. Furthermore, this model may be extended for the purpose of determining the roles and responsibilities of system integrators. / by Allan Fong. / S.M.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2009. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 197-214). / Aerospace systems are among the most complex anthropogenic systems and require large quantities of systems knowledge to design successfully. Within the aerospace industry, an aging workforce places those with the most systems experience near retirement at a time when fewer new programs exist to provide systems experience to the incoming generation of aerospace engineers and leaders. The resulting population will be a set of individuals who by themselves may lack sufficient systems knowledge. It is therefore important to look at teams of aerospace engineers as a new unit of systems knowledge and thinking. By understanding more about how teams engage in collaborative systems thinking (CST), organizations can better determine which types of training and intervention will lead to greater exchanges of systems-level knowledge within teams. Following a broad literature search, the constructs of team traits, technical process, and culture were identified as important for exploring CST. Using the literature and a set of 8 pilot interviews as guidance, 26 case studies (10 full and 16 abbreviated) were conducted to gather empirical data on CST enablers and barriers. These case studies incorporated data from 94 surveys and 65 interviews. From these data, a regression model was developed to identify the five strongest predictors of CST and facilitate validation. Eight additional abbreviated case studies were used to test the model and demonstrate the results are generalizable beyond the initial sample set. To summarize the results, CST teams are differentiable from non-CST teams. / (cont.) Among the most prevalent differentiators is a team's self-reported balance between individual and consensus decision making. Teams that engage in consensus decision making reported stronger engagement in collaborative systems thinking. Another differentiator is the median number of past program experiences on a team. Teams whose members reported more past similar program experiences also reported more engagement in collaborative systems thinking. Data show the number of past similar programs worked is a better predictor than years of industry experience. The apparent enabling effects of qualitative team traits are also discussed. The conclusions of this document propose ways in which these findings may be used to improve training and team intervention within industry, academia, and government. / by Caroline Marie Twomey Lamb. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2010. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 125-126). / The ever-increasing role of intelligence, surveillance, and reconnaissance (ISR) assets in combat may require relatively large numbers of earth observation spacecraft to maintain situational awareness. One way to reduce the cost of such systems is to operate at very low altitudes, thereby minimizing optics size and cost for a given ground resolution. This outside-the-box idea attempts to bridge the gap between high-altitude aerial reconnaissance platforms and traditional LEO satellites. Possible benefits from such a design include enabling a series of cheap, small satellites with improved optical resolution, greater resistance to adversary tracking, and 'quick strike' capability. In this thesis satellite systems design processes and tools are utilized to analyze advanced concepts of low perigee systems and reduce the useful perigee boundary of satellite orbits. The feasibility and utility of such designs are evaluated through the use of the Satellite System Design Tool (SSDT), an integrated approach using models and simulations in MATLAB and Satellite Tool Kit (STK). Finally a potential system design is suggested for a conceptual Continuous Low Orbit Surveillance Satellite (CLOSeSat). The proposed CLOSeSat design utilizes an advanced propulsion system and swooping maneuvers to improve survivability and extend lifetime at operational perigees as low as 160 kilometers, with sustained circular orbits at 240 kilometers. The views expressed in this thesis are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government. / by Jared K. Krueger. / S.M.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2017. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 165-175). / In this thesis, we explore the problem of high-speed autonomous navigation for a dynamic mobile robot in unknown environments. Our objective is to navigate from start to goal in minimum time, given no prior knowledge of the map, nor any explicit knowledge of the environment distribution. Faced with this challenge, most practical receding-horizon navigation methods simply restrict their action choices to the known portions of the map, and ignore the effects that future observations will have on their map knowledge, sacrificing performance as a result. In this thesis, we overcome these limitations by efficiently extending the robot's reasoning into unknown parts of the environment through supervised learning. We predict key contributors to the navigation cost before the relevant portions of the environment have been observed, using training examples from similar planning scenarios of interest. Our first contribution is to develop a model of collision probability to predict the outcomes of actions that extend beyond the perceptual horizon. We use this collision probability model as a data-driven replacement for conventional safety constraints in a receding-horizon planner, resulting in collision-free navigation at speeds up to twice as fast as conventional planners. We make these predictions using a Bayesian approach, leveraging training data for performance in familiar situations, and automatically reverting to safe prior behavior in novel situations for which our model is untrained. Our second contribution is to develop a model of future measurement utility, efficiently enabling information-gathering behaviors that can extend the robot's visibility far into unknown regions of the environment, thereby lengthening the perceptual horizon, resulting in faster navigation even under conventional safety constraints. Our third contribution is to adapt our collision prediction methods to operate on raw camera images, using deep neural networks. By making predictions directly from images, we take advantage of rich appearance-based information well beyond the range to which dense, accurate environment geometry can be reliably estimated. Pairing this neural network with novelty detection and a self-supervised labeling technique, we show that we can deploy our system initially with no training, and it will continually improve with experience and expand the set of environment types with which it is familiar. / by Charles Andrew Richter. / Ph. D.

Massachusetts Institute of Technology. Dept. of Aeronautics and Astronautics. Thesis. 1971. Sc.D. / MICROFICHE COPY ALSO AVAILABLE IN AERO LIBRARY. / Vita. / Bibliography: leaves 441-454. / Sc.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1987. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND AERO. / Bibliography: leaves 227-233. / by Thomas Wesley Roberts. / Ph.D.

Thesis: S.M., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2014. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. "This is joint work with Jana Tumova, Pratik Chaudhari and Sertac Karaman"--Page 3. / Includes bibliographical references (pages 87-89). / We consider the problem of control strategy synthesis for robots given a set of complex mission specifications, such as "eventually visit region A and then return to a base", "periodically survery regions A and B" or "do not enter region D". We focus on problem instances where there does not exist a strategy that satisfies all the specifications, and we aim to nd strategies that satisfy the most important specifications albeit violating the least important ones. We focus on two particular problem formulations, both of which take as input the mission specifications in the form of Linear Temporal Logic (LTL) formulae. In our first formulation we model the robot as a discrete transition system and each of the specifications has a reward associated with its satisfaction. We propose an algorithm for finding the strategy of maximum cumulative reward which has a significantly better computational complexity than that of a brute-force approach. In our second formulation we model the robot as a continuous dynamical system and the specifications are associated with priorities in such a way that a specification with priority i is infinitely more important than one with priority level j, for any i < j. For this purpose, we introduce a functional that quantifies the level of violation of a motion plan and we design an algorithm for asymptotically computing the control strategy of minimum level of violation among all strategies that guide the robot from an initial state to a goal set. For each of our two formulations we demonstrate the usefulness of our algorithms in possible applications through simulations, and in the case of our second formulation we also carry experiments on a real-time autonomous test-bed. / by Luis I. Reyes Castro. / S.M.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2015. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 135-143). / I envision a system that enables successful collaborations between humans and machine learning models by harnessing the relative strength to accomplish what neither can do alone. Machine learning techniques and humans have skills that complement each other - machine learning techniques are good at computation on data at the lowest level of granularity, whereas people are better at abstracting knowledge from their experience, and transferring the knowledge across domains. The goal of this thesis is to develop a framework for human-in-the-loop machine learning that enables people to interact effectively with machine learning models to make better decisions, without requiring in-depth knowledge about machine learning techniques. Many of us interact with machine learning systems everyday. Systems that mine data for product recommendations, for example, are ubiquitous. However these systems compute their output without end-user involvement, and there are typically no life or death consequences in the case the machine learning result is not acceptable to the user. In contrast, domains where decisions can have serious consequences (e.g., emergency response panning, medical decision-making), require the incorporation of human experts' domain knowledge. These systems also must be transparent to earn experts' trust and be adopted in their workflow. The challenge addressed in this thesis is that traditional machine learning systems are not designed to extract domain experts' knowledge from natural workflow, or to provide pathways for the human domain expert to directly interact with the algorithm to interject their knowledge or to better understand the system output. For machine learning systems to make a real-world impact in these important domains, these systems must be able to communicate with highly skilled human experts to leverage their judgment and expertise, and share useful information or patterns from the data. In this thesis, I bridge this gap by building human-in-the-loop machine learning models and systems that compute and communicate machine learning results in ways that are compatible with the human decision-making process, and that can readily incorporate human experts' domain knowledge. I start by building a machine learning model that infers human teams' planning decisions from the structured form of natural language of team meetings. I show that the model can infer a human teams' final plan with 86% accuracy on average. I then design an interpretable machine learning model then "makes sense to humans" by exploring and communicating patterns and structure in data to support human decision-making. Through human subject experiments, I show that this interpretable machine learning model offers statistically significant quantitative improvements in interpretability while preserving clustering performance. Finally, I design a machine learning model that supports transparent interaction with humans without requiring that a user has expert knowledge of machine learning technique. I build a human-in-the-loop machine learning system that incorporates human feedback and communicates its internal states to humans, using an intuitive medium for interaction with the machine learning model. I demonstrate the application of this model for an educational domain in which teachers cluster programming assignments to streamline the grading process. / by Been Kim. / Ph. D.