Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 87-93). / Electricity power systems, typically a very slow-moving and traditional industry, is in a state of flux as technological innovations, such as rooftop solar, home energy management systems, and electric vehicles, are being rapidly integrated into electric distribution systems. As the need to decarbonize the electricity sector becomes increasingly important, a distribution system operator could serve a useful purpose by operating distribution systems and acting as the market operator at a sufficiently granular level to potentially improve resiliency, decrease delivery losses, and send appropriate price signals to its customers. Currently, this latter functionality is assumed to be done using centralized economic dispatch. Given a very large number of small customers and their diverse preferences, it would be computationally expensive to implement centralized economic dispatch at the distribution level with perfect information. In this thesis, an alternative algorithm, referred to as decentralized economic dispatch, is introduced which dispatches power for radial electric distribution systems while accounting for heterogeneous demand functions across customers, demonstrating computationally feasibility, and respecting the physical limits of the system. Unlike other approaches proposed in literature, which often take many iterations or do not converge, the algorithm introduced here converges to the same solution as a centralized operator with perfect information, and does so with only two sweeps across the system. A proof-of-concept example on a 46-bus system demonstrates the physical and economic benefits of the distributed algorithm with varying levels of distributed energy resources. / by Sruthi Davuluri. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 93-96). / An increase in the stock of high skilled workers boosts labor productivity, though economic theory suggests some of the effect may be attenuated by skill mismatch. This research begins to identify and quantify the mechanisms through which skill mismatch affects employment outcomes. Several unique characteristics of personnel management in the US military, particularly in the Air Force, make it an attractive object of study for estimating the magnitude and range of such costs. Unlike most civilian employers, the Department of Defense directly observes the skills of incoming recruits through their achievement on the Armed Forces Qualification Test. Skill requirements for all specialties are expressed in terms of minimum qualifying scores. The matching process produces a differential between a worker's skills and the skills required by her job. Using historical Air Force recruitment and career data, it is possible to identify the relationship between skill mismatch and employment outcomes such as retention and tenure. Using a probit model specification and an instrumental variables approach, this research finds that poorly-matched workers are 20 percent less likely to be retained relative to well-matched workers. Effects of mismatch are most pronounced among high-aptitude workers. The difference in response to skill mismatch between men and women is statistically indistinguishable. These estimates offer a means to quantify the benefits an employer can expect from more thorough evaluation and better job matching of prospective workers. Technological innovations in online learning platforms and skill evaluations offer opportunities to improve matching outcomes. Strategic partnerships between online learning content providers, businesses and workers are important to make these improvements a reality. / by Alexandra Mihaylova. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 74-76). / Over the course of the 2 0 th century numerous economies have leveraged export-driven industrialization strategies to grow wealthier. The advent of automation technology, however, threatens to disrupt the low-cost manufacturing models which have characterized this process; the future may see factories resituated to high-income, high-skill countries which can successfully deploy automation. This thesis consequently evaluates how developing countries could navigate automation by either innovating abreast of it or specializing away from its impact. It is broadly divided into three sections. First, the stage is set by examining the political economy of industrial policy to highlight how political incentives constrain feasible strategies for economic readjustment of any sort. It is shown that even in a setting with few corruption problems - the European Union - industrial policy is guided by politicians' incentives to maintain power, and thus one ought to be cognizant of such incentives in any context. Second, possible barriers to greater productivity and innovation in developing countries are explored through a case study analysis of Vietnam, which is considered by some to be highly exposed to automation risk. Growth diagnostic tools are applied to identify the binding constraints which prevent it from shifting towards more complex, value-added economic activities. Structural economic reform is found to be critical to greater innovation, as opposed to technocentric solutions that aim to leapfrog to the technological frontier. Third, product space and machine learning methodology are used to simulate how countries' export diversification paths could respond to automation. By conducting sensitivity analysis across a range of automation scenarios it provides insight on how developing countries may be able to respecialize their economies to maintain growth. / by Eric Protzer. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 27-30). / The development of self-propagating gene drive containing organisms may have wide-reaching consequences, including for international trade. Given compelling potential applications in disease mitigation and pest management, countries may develop and release these self-propagating products of biotechnology in the near future. Such products will cross international borders through natural organism movement and through trade, potentially altering not only the populations and ecosystems of the country that developed the product, but also those of the countries to which they spread. Political and economic consequences will arise, some of which the World Trade Organization (WTO) will govern. The WTO's Agreement on Sanitary and Phytosanitary Measures (SPS Agreement) is an especially pertinent set of trade rules which permits trade restrictions aimed at protecting the health of a country's citizens or other species only when scientific evidence supports the perception of risk to the health of citizens or other species. In the early 2000s, the WTO evaluated the SPS Agreement for applicability to genetically modified crops, but no such analysis has yet been completed for gene drive containing organisms. In this thesis, I first summarize the status of gene drive development, highlighting divergent techniques and their status in laboratory experiments. I then explore potential scenarios for trade disputes involving gene drive containing organisms, with differing technological and ecological specifications. I examine the text of the SPS Agreement, as well as its past application to trade disputes, to evaluate relevance to gene drive containing organisms. Finally, I formulate suggestions to policy makers and technology developers for preventing international disputes on, and disruption of trade by, gene drive containing organisms. / by Lily Fitzgerald. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages [58]-61). / This thesis assesses electricity distribution network investment decision-making methods using a three-step approach, to explore the possible value of flexibility. By applying a simple quantitative framework to an illustrative distribution network decision, it finds that current methods that fail to consider flexibility can result in higher cost investments and lost value. First, it presents the recent developments in the electricity sector and outlines how the current state of network planning is no longer fit for purpose. Second, it proposes a flexible design approach for electricity network investments, and determines the value of this flexibility by developing a simple model and applying real options analysis. Third, it identifies the practical challenges to effectively implementing a flexible design methodology, before proposing recommendations for electricity utilities and regulators. The proposed flexible design methodology found that building in flexibility through Non-Wire Alternatives provided a greater Expected Net Present Value than current robust techniques using traditional investments. This thesis confirmed that the value of this flexibility increased as uncertainty over future electricity demand increased. This thesis finds there is a strong case that the use of a flexible design approach can increase the cost-effectiveness of network investment decisions. However, there remains significant uncertainty regarding key parameters that determine the value of these cost savings. As such, this thesis concludes with a discussion of evidence gaps and priorities for future research. / by Anna Evans. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, Technology and Policy Program, September, 2020 / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 81-86). / MIT Lincoln Laboratory is developing a software-reconfigurable imaging architecture called ReImagine, the first field programmable imaging array (FPIA), which reflects a broader trend in the increased use of FPGAs in sensor systems in order to reduce size and power consumption without a corresponding loss in performance or flexibility. At the same time, the field of machine learning is diversifying to include distributed deep learning methods like split learning, which can help preserve privacy by avoiding the sharing of raw data and model details. In order to continue to expand the capabilities of architectures like ReImagine's and enable split learning and related techniques to be used in the growing body of FPGA-based sensor systems, we examine the relationship of emerging split learning applications to FPGA-based image processing platforms. We determine that the implementation of split learning methods on FPGAs is feasible, and outline use cases in the areas of health and short timescale physics that demonstrate the usefulness of these implementations to both organizations concerned with privacy-preserving machine learning methods and organizations concerned with the deployment of efficient, flexible, and low-latency sensor systems. We begin by conducting a survey of the modern FPGA landscape in terms of technical attributes, use in sensor systems, security and privacy features, and current machine learning implementations. We also provide an overview of split learning and other distributed deep learning methods. Next, we synthesize an example split learning model in HDL code to demonstrate the feasibility of implementing such a model on an FPGA. Finally, we develop use cases for split learning applications on FPGA-based sensor systems and offer conclusions about the future development of distributed deep learning on heterogeneous processing platforms. / by Hannah K. Whisnant. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, Technology and Policy Program

This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from student-submitted PDF version of thesis. Vita. / Includes bibliographical references (pages [79]-82). / The study of social systems has increasingly relied on data collection and analysis to draw conclusions. In parallel, the research community has often tried to understand entrepreneurs quantitatively, e.g. by understanding which behaviours or personality traits most often correlated with entrepreneurial success. While a quantified representation is essential in modeling what is being studied, it hides away the mental processes that create behaviours. Arguably, the people who engage the most directly with systemic change are entrepreneurs. They have to create their own system (a company) and connect it to its wider network (clients, investors, etc.). Most importantly, the more uncommon their idea is, the less they can rely on existing frames to bring their ideas to fruition, and the more they have to reflect on the dynamics of their wider context and how their company can integrate to it. / According to constructive-developmental theory (CDT), the ability to see context dynamically is not a personality trait, but a structure of perception. Our meaning-making--how we create that perception of the world--grows in complexity as we delve into our reflections. CDT makes the separation between a structure in which value is defined by the contexts that we are in, and the next more complex one, where the self defines value. This thesis is an early attempt at understanding the experience of systemic change, and the growth in perception that happens along-side it. By bringing CDT and associated theories of the mind into the experience of entrepreneurship, I sought to understand how entrepreneurs make meaning of themselves in the course of their growth, by asking them directly. This is the first known study that maps the meaning-making complexity of entrepreneurs using the subject-object interview, CDT's qualitative empirical research method. / I found that, as perception became more intrinsic, the definition of value went from seeking to have what we wish to have, to leaving a legacy through impact, to acting in harmony with our meaning of value. Instead of seeking purpose, entrepreneurs who perceive intrinsically act on the one they already have. Instead of seeking value, they share it with others in all their interactions. / by Othmane Benkirane. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 91-93). / The migration of news to the web has given advertisers new opportunities to target readers with ever more personal and engaging ads. This sponsored content, known as native advertising, is placed in news publications often camouflaged as legitimate news. Though native ads bring revenue to the struggling U.S. news industry, their ability to draw loyal readers off-site could hurt publishers in the long run. Herein, I measure the quality and the impact of the ads from Content Recommendation Networks (CRN) on the U.S. news industry, between March 2016 and February 2019. A CRN controls both the third-party ads and the house ads -- recommendations for news articles from the host publisher -- on a news publisher's website. During the 2016 presidential election, I found that 17% of ad headlines were political, and 67% of the stories were clickbait. Over the 2018 midterm elections, 15% of the ads were political, and 73% were clickbait. While third-party ads are more clickbait than house ads, the increase in clickbait between 2016 and 2018 is larger for the house ads. Further, I investigate the effect that a one-time exposure to these ads have on the perceived credibility on news articles. Four publishers were under study: CNN, Fox News, The Atlantic and Sacramento Bee. A one-time exposure to CRN ads was found to have no signicant eect on the credibility of traditional publishers. Yet, the CRN ads impacted the credibility of less well-known publishers: ads increased the credibility of the news on Sacramento Bee, and decreased it on The Atlantic. / "Funded by the Vannevar Bush Faculty Fellowship from the Department of Defense awarded to Professor Jadbabaie"--Page 7 / by Manon Revel. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 145-151). / Uncertainty and risk are endemic in research and development (R&D) programs. Managing risks is a pivotal step at all phases of system design, but is particularly important at the beginning phases where system designers attempt to assess many design alternatives before the systems is developed and deployed. Designing systems that mitigate risks or are flexible enough to withstand them can lead to more reliable and valuable systems. If decision-makers believe the risks are too high, they may not invest in a technology that would provide the capabilities they need. Alternatively, if they underestimate the risks of a technology, investment may lead to costly and embarrassing overruns. Unfortunately, the existing approaches that decision-makers use to help them understand their options are often inadequate and do not tell the full story. / In practice, overly simplistic methods can lead to arbitrary analysis, while more complex techniques such as NASA's probabilistic risk assessment procedure can be too costly and time-consuming for the fast-moving R&D context. These shortfalls can lead to results that decision-makers do not trust and ultimately ignore, leading to decisions based on gut-feelings or hunches. In addition to these shortfalls, there exist gaps between the theory upon which many decision methods and their results are based; several researchers have pointed out computational weaknesses in risk matrices, one of the most widely-employed risk assessment tools. Simply put, the decision-making tools that exist today may not be adequate. Moreover, many risk assessments are conducted secondary to initial tradeoff analysis, resulting in suboptimal design decisions that do not account for risks from the beginning. / In R&D programs, decision-makers require a human-centric method of modeling and communicating the benefits, costs, and risks associated with various technologies. This research develops a framework for conducting Visualized Risk-Informed Tradespace Exploration (vRITE) based on insights gained from practicing R&D program managers. vRITE provides decision-makers an interactive, human-centric analysis of the cost, schedule, performance, and risk associated with multiple technologies. Two example vRITE analyses are used illustrate how the framework addresses the real-world objective and subjective considerations that decision-makers care about most and allows them to explore the data, helping them identify which technology warrants investment. The information provided in a vRITE analysis can reduce the barriers to R&D investment and ultimately lead to more risk-informed decision-making. / by Samuel Adam Mahle. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 77-84). / We discuss a problem: Internet of Things devices running software are vulnerable to accidents and exploitation, a technology solution: preventing exploitable bugs by developing machine-checked proofs of software correctness and security, and a policy lever to incentivize adoption of this solution: a safe harbor from FTC unfairness prosecution for manufacturers that use formal methods to guarantee safer, more secure devices. To motivate the potential of formal methods, we present a technical contribution: a formally verified connected lightbulb switch, proven immune to certain types of software exploits. We discuss a framework, the Common Weakness Enumeration, that the FTC and manufacturers could use as a shared language to explain what classes of software vulnerability a manufacturer will defend against. We outline the authority of the FTC in regards to poor data security practices as unfair practices and how our safe harbor would both provide immunity to participants and be updated over time to continue to incentivize ever stronger software protections. / by Clark Wood. / S.M. in Technology and Policy / S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society