Title
stringlengths 3
331
| text
stringlengths 14
9.14k
⌀ |
|---|---|
Innovative approach to understanding & addressing obesity In the Kingdom of Saudi Arabia | Obesity and overweight are complex global issues that have been and continue to be a significant problem that needs to be addressed. Understanding obesity and overweight are fundamental to finding practical and sustainable solutions. Innovation has different meanings to different people and can be applied in many different sectors in varying forms and at different levels. Innovation in healthcare is no longer a luxury but rather a necessity. In this thesis, we review some concepts of interest to healthcare innovation briefly and also examine the topic of obesity and overweight from a global perspective and with a focus on the Kingdom of Saudi Arabia. We review global obesity and overweight and then focus on obesity and overweight in the Kingdom of Saudi Arabia. We also reflect on the work of Prof. Clayton Christensen "Jobs to be Done Theory" and how it can help address innovation in the healthcare system and in particular applying it to the general concept of tackling obesity. The thesis highlights a critical understanding of obesity based mostly on the work of Jason Fung, MD in his recent book titled "The Obesity Code: Unlocking the Secrets of Weight Loss." A novel integrated solution for tackling obesity in the Kingdom of Saudi Arabia will be proposed incorporating insights from the research material on both innovation and obesity with the utilization of concepts gained from the System Design and Management program at MIT. The concept of innovation in the healthcare setting is shown to be instrumental in creating an opportunity for higher quality, cheaper and faster delivery of health services heavily dependent on the work of Prof. Clayton Christensen. Also, we highlight the need for an innovative integrated solution at different levels of the system including the individual level and institute level and finally the national levels. We believe there is a genuine need to approach innovation in the healthcare setting at the different establishments within the healthcare system and the importance of cross-pollinated innovation teams. |
The planning and real estate implications of North American retirees moving to Mexico : A study of Bucerías, Nayarit | Over the past decades, thousands of U.S. and Canadian retirees have moved to Mexico in search of better living conditions. Although the total number of foreign retirees in Mexico represents a small fraction of the total Mexican population, their concentration in specific communities suggests that the impact on these communities may be significant. Furthermore, a series of circumstances such as the North American Free Trade Agreement, the Internet, as well as the imminent demographic changes in the U.S. population all suggest that a dramatic increase in the number of retirees moving to Mexico might occur. Recent trends among the North American retirees that are moving to Mexico suggest that the retirees are becoming younger and that they are increasingly locating in Mexico's northwestern coasts. While the impact that the presence of North Americans may have in small coastal communities may be important, there is a lack of research on this phenomenon. This thesis is an exploratory study of a particular community, Bucerías (Nayarit), which is experiencing this phenomenon. The thesis intends to unveil the major planning and real estate implications of foreign retirees living in one such coastal community. The study is mostly based on extensive qualitative on-site interviews. Useful evidence for both local authorities as well as for real estate and tourism industry professionals on the positive and negative implications of the presence of retirees is provided. |
Computational prediction of ribonucleic acid-based gene regulatory mechanisms in human and Tetrahymena | The diversity and profound impact of gene regulation mediated by small RNAs (sRNAs) is just beginning to come into focus. RNA interference (RNAi) pathways have been shown to mediate processes such as genomic rearrangement in ciliates and developmental timing and tissue differentiation in plants and animals. Here we present a computational study into the function of two distinct classes of sRNAs. In the first section, we examine an uncharacterized class of sRNAs isolated from the ciliate Tetrahymena thermophila, present functional comparison to known classes of sRNAs in other organisms, and note a strong and specific relationship to a novel sequence motif. In the second section, we examine the evolutionary impact of microRNAs (miRNAs), which mediate potent post-transcriptional repression on their targets. We observe that miRNAs with tissue-specific expression exert remarkable evolutionary pressure, compelling many preferentially coexpressed genes to avoid accumulating target sites. We present tissue-specific patterns of such target depletion and note strong agreement with experimentally obtained miRNA expression patterns. Conversely, we report enrichment for targeting among genes with expression patterns spatially or temporally complementary to the miRNAs', suggesting a widespread role of tissue identity maintenance for miRNA-mediated regulation. |
Oxygen defect engineering for silicon solar cell applications : process design and modeling | One of the main remaining impurities that lowers efficiencies of silicon solar cells are oxygen ring defects that are incorporated into the material during growth. These defects decrease overall cell efficiencies by around 20% (rel.) resulting in a yield loss of about 1/4 h of each monocrystalline silicon ingot. To control the oxygen defects and put them in the least harmful form possible, a new cell processing step call tabula rasa (TR) is explored. TR is a high temperature process for a short duration. In this work, TR is found to be a kinetically-limited process through several oxygen precipitate dissolution experiments from which the activation energy of dissolution is found to be equivalent to the migration enthalpy for oxygen in silicon. With this knowledge, a predictive kinetic model is built which can be used for process optimization. A multiscale end-to-end model is also developed to determine the effect of rings on cell performance. Using oxygen defect parameter inputs, device modelling, and a spatially resolved two diode mesh, PL images are transformed into current maps and used to determine cell efficiencies for inhomogeneously distributed defects. A reduction in efficiency for cells with ring defects is simulated for several ring defect concentrations and compared to a non-defective cell. Another strategy for lowering cost and mitigating oxygen ring defects is using thinner silicon absorbers with inherent defect tolerance. Using the multiscale modeling platform described above, thin silicon wafers are compared to typical cells. It is shown that thin cells with ring defects perform with higher efficiencies, with less than half the material used. The thin silicon strategy is compared with the TR process addition, and future work is outlined to further explore these oxygen mitigation options. |
On the warm bias along the South-West African Coast in coupled models : an oceanic perspective | Coupled ocean/atmosphere simulations exhibit systematic warm biases over the South West African (SWA) coastal region. Recent investigations indicate that coastal ocean dynamics may play an important role in determining the SST patterns, but none of them provide a detailed analysis. In this study, I analyze simulations produced both by coupled models and by idealized models. Then results are interpreted on the basis of a theoretical framework. Finally the conclusion is reached that the insufficient resolution of the ocean component in the coupled model is responsible for the warm biases over the SWA coastal region. The coarse resolution used in the ocean model has an artificially stretched coastal side-wall boundary layer, which induces a smaller upwelling velocity in the boundary layer. The vertical heat transport decreases even when the volume transport is unchanged because of its nonlinear relationship with the magnitude of the upwelling velocity. Based on the scaling of the idealized model simulations, a simplified calculation shows that the vertical heat transport is inversely proportional to the zonal resolution over the coastal region. Therefore, increasing the horizontal resolution can considerably improve the coastal SST simulation, and better resolve the coastal dynamics. |
A priori and on-line route optimization for unmanned underwater vehicles | The U.S. military considers Unmanned Underwater Vehicles (UUVs) a critical component of the future for two primary reasons - they are effective force multipliers and a significant risk-reducing agent. As the military's technology improves and UUVs become a reliable mission asset, the vehicle's ability to make intelligent decisions will be crucial to future operations. The thesis develops various algorithms to solve the UUV Mission-Planning Problem (UUVMPP), where the UUV must choose which tasks to perform in which sequence in a stochastic mission environment. The objective is to find the most profitable way to execute tasks with restrictions of total mission time, energy, time-restricted areas, and weather conditions. Since the UUV accumulates navigation error over time while maneuvering underwater, the UUV must occasionally halt operations to re-orient itself via a navigation fix. While a navigation fix takes time and increases the likelihood of exposing the vehicle's position to potential adversaries, a reduction in navigation error allows the UUV to perform tasks and navigate with a greater amount of certainty. The algorithms presented in this thesis successfully incorporate navigation fixes into the mission-planning process. The thesis considers Mixed-Integer Programming, Exact Dynamic Programming, and an Approximate Dynamic Programming technique known as Rollout to determine the optimal a priori route that meets operational constraints with a specified probability. The thesis then shows how these formulations can solve and re-solve the UUVMPP on-line. In particular, the Rollout Algorithm finds task route solutions on average 96% of the optimal solution a priori and 98% of the optimal solution on-line compared to exact algorithms; with a significant reduction in computation run time, the Rollout Algorithm permits the solving of increasingly complex mission scenarios. |
Mechanical stiffness-defined matrices for stem cell research and drug screening | Synthetic polymer matrices or subtrata with tailored elastic properties provide a powerful method to direct biological cell' differentiation and foster cell multiplication. By changing the stiffness of the substrate, human mesenchymal stem cell (MSCs) could be directed along neuronal, muscle, or bone lineages. Matrix elastic modulus can also control anchorage dependent cell's motility, localization, tissue formation and organization. Besides that, synthetic materials such as biodegradable polymers offer a versatile alternative to naturally derived biopolymers. Their mechanical properties can be highly tailored and they are easy to synthesize and shape. Moreover, these platforms can be readily "biologically" fine-tuned toward a particular cell linage by incorporating well-documented parameters, which play crucial roles in cell-extra cellular matrix (ECM) signaling pathway, such as growth factor, surface topology and stimulation signal. Hence, these materials are suitable candidates to develop engineered matrices for stem cell culture, cell manipulating platforms in biological research and drug development. In this thesis, commercialization aspects of these engineered matrices for stem cell research, cell culture and drug development markets are evaluated both in USA and in Singapore markets. Technological barriers, intellectual property and a preliminary cost model are analyzed. A business plan is presented and discussed for applications in both the stem cell research and the drug screening markets. Although these two markets are ill-defined, both of them are growing rapidly and appear to be very promising. A review of the technology itself led to the conclusion that the matrix is capable of induce anchorage dependent cell into specific lineage but the success rate is not yet quantified and further research need to be done to achieve good reproducibility and to meet the required efficacy of the industry. |
Analysis of the optical coupling of wavelength-shifting fibers to organic liquid scintillator filled fluoropolymer tubes for industrial and nuclear security applications | Industrial and nuclear security applications continue to push radiation detection development into new and exciting frontiers. In this work, an innovative detection module is developed and tested for use in a cosmic ray imaging (CRI) system designed for oil field characterization and is evaluated for its potential use in a fast neutron detection system for nuclear security applications. By measuring density changes in the reservoir, the CRI system will provide real-time information about steam chamber development during the enhanced oil field recovery process known as steam assisted gravity drainage (SAGD). The ability to monitor the development of the steam chamber region has the potential to provide important information, which could be used to optimize the growth and uniformity of the underground steam chamber and minimize costs. The organic liquid scintillator based detection modules also detect fast neutrons. During the initial characterization of an unidentified radioactive source, it is important to have the capability to determine if special nuclear material (SNM) is present and if it is configured to produce a nuclear yield. The emission of multiple neutrons during a single fission makes it possible to use this unique timing characteristic to identify SNM. The number of specialists trained to handle nuclear devices is limited making this determination a critical step in properly responding to the situation. The detector module consists of a 5 mm diameter by 2-meter long fluoropolymer tube filled with organic liquid scintillator (OLS), optically read-out using wavelength shifting (WLS) fibers. The 1:400 ratio of diameter to length makes light collection from the organic scintillator very challenging. Over ten configurations of OLS, fluoropolymer tubes, and WLS fibers were tested. The final configuration consisted of two 2mm BCF-91A WLS fibers optically coupled to the outside of an optically transparent fluorinated ethylene propylene (FEP) tube filled with a commercial OLS (EJ-309). Cosmic ray muons produce large light pulses in the OLS of which a portion reaches the external WLS fibers. The WLS fibers re-emits the light at longer wavelengths and acts as a multi-mode light guide channeling the signal to photomultiplier tubes located at each end of the WLS fibers. This module demonstrated excellent detection efficiency with less than 5% signal reduction, at any point along the module, due to optical attenuation. Timing analysis of the WLS fiber signals also provided coarse position determination, 40 cm, which opens design options not previously available. An important characteristic required of neutron detectors for nuclear security applications is the ability to discriminate fast neutron and gamma ray events. Initial tests have demonstrated the capability of our module to discriminate neutron and gamma rays by applying the rise time pulse shape discrimination (PSD) method to the WLS fiber signals. EJ-309 is well known for its PSD capabilities. Coupling this desirable characteristic with loss free, low attenuation optical read-out through a WLS fiber has the potential to broaden significantly liquid scintillator applications. |
Value stream financial modeling for improved production decision making | Understanding the overall impact of a decision in a manufacturing system can be challenging given the complex production and financial structures in today's companies. While knowing the direct result of a local change may be easy, anticipating the real impact to the rest of the business can be difficult. Nonetheless, managers are faced with this dilemma on a regular basis as they try to support the larger organization, taking appropriate actions as best they can. Based on a project at Novartis Vaccines and Diagnostics for an influenza vaccine, this thesis helps address some of the key questions managers face. It discusses a technique for more accurately determining the implications of these common manufacturing decisions: * How much should be spent to improve a particular component? * What are the impacts of expanding into new markets? " Which parameters in the factory most deserve managerial attention? * What are the appropriate tradeoffs to make when deciding on materials purchasing? Using concepts from throughput accounting, a model is developed from a detailed cost structure analysis, linking the financial and production aspects of the system. Whenever a parameter is changed, the model simulates how the rest of the system would perform through a linear program that replicates the production scheduling process. Thus, a manager is able to experiment with the tool in order to observe the overall impact of the change being considered and levy a decision based on the anticipated costs and benefits projected by the model. As a result, managers can distribute resources in a more efficient manner and align decision making throughout the organization. This thesis discusses the modeling approach, historical validation and initial insights for the current system. It also covers techniques for future applications and identifies the underlying organizational challenges that must be addressed to achieve a global optimum. |
Data analysis to understand coordination and topological environments in oxides | Local coordination and topology of ions determine several important properties of materials, including electronic structure, migration barrier, and diffusivity. In this thesis, we employ the Materials Project Database to investigate the coordination preferences of cations and topology of coordination polyhedra in oxides. We calculate the coordination environment preferences of several common cations in oxides, identifying lithium, sodium, calcium, and magnesium ion's preferred coordination numbers are 4- fold/6-fold, 6-fold, 6-fold, and 6-/4-fold coordination respectively. We also develop a method to quantify the connectivity between two polyhedra and determine whether they are point-sharing, edge-sharing, or face-sharing. We find that 4-fold coordinated lithium polyhedra mainly point-share while the 6-fold coordinated lithium polyhedra connectivites are face-sharing. We then build a tool to identify and insert "empty polyhedra" (i.e. coordination polyhedra which are bounded by ions but contain no central ion) which can help to provide a better descriptor of the structure topology. We also find that most connections with lithium polyhedra are with empty polyhedra and that in a connected set of two lithium polyhedra and one empty polyhedron, the coordination polyhedra tend to be either 6-4-6 or 4-6-4 with the empty polyhedron in the center. Finally, we utilize the database to evaluate Pauling's first and second rules, which are guidelines for current understanding of coordination and topology, and observe that the rules are generally accurate only within a 30% error margin. |
Bioinorganic hydrocarbon oxidation : mechanistic and kinetic studies of the soluble methane monooxygenase from Methylococcus capsulates (bath) | Chapter 1. Principles of Small Molecule Activation by Metalloproteins as Exemplified by the Soluble Methane Monooxygenase -- Chapter 2. Small Molecule Binding to the Mixed-Valent Diiron Center of Methane Monooxygenase Hydroxylase from Methylococcus capsulatus (Bath) as Revealed by ENDOR Spectroscopy -- Chapter 3. An EPR Study of the Dinuclear Iron Site in the Soluble Methane Monooxygenase Reduced by One Electron at 77 K: the Effect of Component Interactions and the Binding of Small Molecules to the Dinuclear Ferric Center -- Chapter 4. An Investigation of the Reaction of Diferrous Methane Monooxygenase Hydroxylase with Dioxygen and Substrates by Rapid Freeze- Quench and Stopped-Flow Spectroscopy -- Chapter 5. Oxidation of Radical Clock Substrate Probes by the Soluble Methane Monooxygenase System -- Chapter 6. Tritiated Chiral Alkanes as Probes for the Mechanism of Hydroxylation by the Soluble Methane Monooxygenase. |
Carbon dioxide capture from coal-fired power plants : a real potions analysis | Investments in three coal-fired power generation technologies are valued using the "real options" valuation methodology in an uncertain carbon dioxide (CO2) price environment. The technologies evaluated are pulverized coal (PC), integrated coal gasification combined cycle (baseline IGCC), and IGCC with pre-investments that make future retrofit for CO2 capture less expensive (pre-investment IGCC). All coal-fired power plants can be retrofitted to capture CO2 and can be considered "capture-capable", even though the cost and technical difficulty to retrofit may vary greatly. However, initial design and investment that take into consideration such future retrofit, makes the transition easier and less expensive to accomplish. Plants that have such an initial design can be considered to be "capture-ready". Pre-investment IGCC can be considered to be "capture-ready" in comparison to PC and baseline IGCC on this basis. Furthermore, baseline IGCC could be taken as "capture-ready" in comparison to PC. Cash flow models for specific cases of these three technologies were developed based on literature studies. The problem was formulated such that CO2 price is the only uncertain cash flow variable. All cases were designed to have a constant net electric output before and after CO2 retrofit. As a result, electricity price uncertainty had no differential impact on the competitive positions of the different technologies. While coal price was taken to be constant, sensitivity analysis were conducted to show the impact of varying coal prices. Investment valuation was done using the "real options" approach. |
Micropower digital signal processing for sensor applications | Ultra-low power systems, such as wireless microsensor networks or implanted medical devices, are driving the development of processors capable of performing increasingly complicated computations using mere microwatts of power. This thesis describes the design of a micropower DSP intended for medium bandwidth microsensor applications (such as acoustic sensing and tracking) which achieves 4 MIPS performance at 40 [mu]W (10 pJ per instruction) operating at 450 mV and fabricated in 90 nm CMOS. Energy efficiency optimizations include a custom CPU instruction set, a miniature instruction cache with a novel replacement strategy, hardware accelerator cores for FIR filter and FFT operations, and extensive power gating of both logic and memory. The tradeoffs of cache size, line length, and replacement policy for very small (a few hundred words or less) caches are explored, as are the design implications of optimizing the cache for minimum energy without regard to performance (since onchip memory access is already single-cycle). A replacement policy designed to reduce thrashing in miniature instruction caches is presented. Efficient control of power-gated circuits requires consideration of the minimum off time, or break-even time. An energy model for determining the break-even time is developed, which correlates with measurements of the power-gated domains on the DSP. The energy savings obtained from hardware accelerators for FIR filtering and FFT operations are measured, and a model is developed to predict the actual net power reduction in a real system, including factors such as sampling rate, leakage power, latency requirements, and power gating overhead. |
Emerging technology intelligence : scanning and monitoring for strategic planning | In recent years we have seen a confluence of science and technology that portends great impact to business strategy and product development. Just as physicists, chemists and biologists have extended their knowledge to other fields in pursuit of their own research, so, too, must senior business managers extend beyond competency in finance and marketing to comprehend fundamental scientific principles in order to make strategic business decisions. In addition to technical knowledge specific to their particular industry, successful leaders of the very near future will have to understand the ebb and flow of new technology discoveries "emerging" from the lab to the market, and will skillfully leverage these dynamics to navigate their companies through waves of innovation. The speed of scientific discovery and the convergence of multiple technologies to influence a single product create a compelling argument for corporate investigation of a range of emerging technologies and tracking of developments in multiple industries. "Technical intelligence" is knowledge, derived from analysis of appropriate metrics, trends and activities, which informs strategic planning and decision-making. This research surveys the academic literature and examines some methods and tools for gathering and employing technical intelligence. We conclude that an amalgamation of strategic evaluation methods, but especially scanning and monitoring, can and should be used to effectively develop an objective, simple, and descriptive view of technology emergence that captures activity, momentum, and ultimately, viability, of emerging technology over time. |
Retrofit systems for reconfiguration in civil aviation | A new concept for retrofitting a reconfiguration module to an existing control law is reported in this thesis. The concept is motivated by the need for low cost, add-on modules that improve air safety in the existing fleet of civil air transport vehicles. A direct adaptive approach that accommodates control surface nonlinearities is adopted, which uses a slowly adapting model of the closed-loop aircraft as the reference model. The motivation, benefits, and components of the architecture are presented. In addition, the issues of control surface magnitude and rate saturation are addressed. A proof of stability is outlined for input-error adaptation when position and rate saturation are present. The reconfiguration architecture is demonstrated using an F/A-18 and a generic transport nonlinear simulator. General issues associated with commercial transport reconfiguration are highlighted. In both the longitudinal and directional axes, the control surfaces are not well balanced from a reconfiguration viewpoint. As a result, a novel reconfiguration control allocation scheme was devised that blends in all the control effectors in a given axis to perform the reconfiguration task. The simulation results revealed that the reconfiguration architecture does provide reconfiguration functionality for a wide variety of control surface failures. The reconfiguration potential is illustrated through comparisons of post-failure performance with and without reconfiguration via non-linear simulations. Additionally, comparisons between post-failure performance and nominal performance are made through non-linear simulations, closed-loop frequency responses, and aircraft handling qualities. For all of the failure scenarios illustrated, the simulation results showed that the aircraft without reconfiguration departs; with reconfiguration, nominal performance is achieve provided that adequate control authority exists post-failure. |
Correlated exciton dynamics in semiconductor nanostructures | The absorption and dissipation of energy in semiconductor nanostructures are often determined by excited electron dynamics. In semiconductors, one fundamentally important electronic state is an exciton, an excited electron bound to a positively charged vacancy. Excitons can become correlated with other excitons, mutually influencing one another and exhibiting collective properties. The focus of this dissertation concerns the origins, effects, and control of correlated excitons in semiconductor nanostructures. Correlated Coulomb interactions can occur between excitons, resulting in energy shifts and dephasing in each exciton. Two-dimensional Fourier-transform optical spectroscopy is a powerful tool to understand Coulomb correlations; the technique relates exciton dynamics during distinct time periods. However, the technique is still limited by weak spectral features. Using two-dimensional pulse shaping methods, waveforms of excitation fields were tailored to selectively amplify spectral features of correlated exciton states in gallium arsenide quantum wells. With the aid of theoretical models, 2D spectra of quantum wells revealed clear contributions of Coulomb correlations to the exciton dynamics. Time and power dependent properties of the 2D spectra indicate several mechanisms for exciton interactions that are neglected in commonly used theoretical models. If a semiconductor material is fabricated within a microcavity, optical fields can be trapped around the semiconductor, strongly distorting properties of the semiconductor excitons and forming new quasi-particles called exciton-polaritons. Theoretical work has suggested exciton-polariton Coulomb correlation strengths can be reduced compared to that of excitons. Using two-dimensional Fourier-transform optical spectroscopy, control of Coulomb correlations was demonstrated by varying the cavity structure. The cavity fields were also shown to induce high-order correlated interactions among exciton-polaritons. A macroscopic quantum degenerate system of exciton-polaritons can also become correlated, exhibiting long-range order typical of a Bose-Einstein condensate. However, unlike a Bose-Einstein condensate, exciton-polartions are not typically in thermal equilibrium. Using a sample with exciton-polariton lifetimes longer than previous samples, the macroscopic behavior of exciton-polaritons was investigated by imaging the exciton-polariton photoluminescence. Condensation depended significantly on spatially-varying potential energy surfaces. Using optically-induced harmonic potential barriers, thermal equilibrium among exciton-polaritons was achieved, with exciton-polaritons forming a Bose-Einstein distribution at densities above and below the condensation phase transition. |
How a utility and its street-level bureaucrats connected the poor in Bangalore | This dissertation is about how urban water utilities behave and what makes them interested in serving the poor. The infrastructure literature tends to treat public service agencies as monolithic entities and to ignore the great diversity of tasks and behavior patterns within them. As a consequence, common explanations for why utilities fail poor people tend to focus on attributes of the external environment in which utilities sit and not on the potential to elicit interest from within. This research corrects for this bias by applying a "street-level bureaucracy" approach to a study of a large urban water utility. The aim is to quash the notion so common in the water literature of a unified agency operating on the supply side and to rekindle an interest in the actions of workers. To do this, I examine the case of the Bangalore Water Supply and Sewerage Board (BWSSB) and its contrasting outcomes within the same case. Over a five year period from 2000 to 2005, the utility revised its operational policies to accommodate the legal and financial realities of slums and connected 5,000 households or five percent of the slum population to the water network. Although the BWSSB demonstrated an unusual commitment to the poor, its efforts were not an unmitigated success. Progress was slow and staff failed to connect households to the network in many of the slums targeted. This dissertation digs deep inside the utility to explain these contrasting outcomes holding the city, the agency, and the sector efficiency constant. I find that while external pressures were necessary to prompt a business-as-usual utility to take action in slums, variation in outcome can be explained by the different facets of engineering life in BWSSB service stations and the different kinds of relationships forged between frontline staff and slum dwellers. |
Mapping comfort : an analysis method for understanding diversity in the thermal environment | Our thermal experience is never neutral. Whether standing near a cold window in the winter, or in the shade on a sunny day, we constantly experience a rich set of thermal stimuli. Yet, many of the tools used in professional practice to analyze and design thermal environments in buildings do not account for the richness of our thermal experience. This disconnect between our analysis tools and our experience results in buildings that use more energy than they should, and that leave occupants dissatisfied with their thermal environment. This thesis seeks to bridge the gap between our thermal experience and our building thermal analysis tools. A unique methodology has been developed that produces mapping of thermal comfort parameters in all three spatial dimensions, as well as over time. Both heat balance and adaptive comfort indices have been incorporated into the methodology. An accompanying software program, called cMap, has been developed to illustrate the ways that this methodology can be used with existing energy analysis software and to demonstrate how it can fit into existing analysis workflows in professional practice. |
Advanced photoanodes for photoassisted water electrolysis | With continuously growing energy demands, alternative, emission-free solar energy solutions become ever more attractive. However, to achieve sustainability, efficient conversion and storage of solar energy is imperative. Photoelectrolysis harnesses solar energy to evolve hydrogen and oxygen from water, thereby enabling energy storage via chemical means. Hematite or [alpha]-Fe₂O₃ has emerged as a highly promising photoanode candidate for photoelectrochemical cells. While significant improvements in its performance have recently been achieved, it remains unclear why the maximum photocurrents still remain well below their theoretical predictions. This study investigates the defect chemistry and conduction mechanism of hematite in order to understand and improve this material's shortcomings. A defect model for donor doped hematite was derived and its predictions conformed by the electrical conductivity of ilmenite hematite solid solution bulk samples as a function of temperature and oxygen partial pressure. The enthalpies of the Schottky defect formation and the reduction reaction for hematite were determined as 13.4 eV and 5.4 eV, respectively. In addition, a temperature independent value for the electron mobility of 0.10 cm2/Vs for 1% Ti donor doped hematite was derived. Furthermore, the electrical conductivity of nanometer scale, epitaxially grown thin films of the ilmenite hematite solid solution system was characterized by electrical impedance spectroscopy. This work reports a detailed correlation between the electrical conductivity of the undoped hematite, the 1 atom% Ti doped hematite and the thin films with higher ilmenite content and the conditions under which they were annealed (20° C=/< T =/< 800° c and 10-4 atm =/< po2 =/< atm). Hematite's room temperature conductivity can be increased from ~10-11 S/cm for undoped hematite films by as much as nine orders of magnitude by doping with the Ti donor. Furthermore, by controlling the non-stoichiometry of Ti-doped hematite, one can tune its conductivity by up to five orders of magnitude. Depending on processing conditions, donor dopants in hematite may be compensated largely by electrons or by ionic defects (Fe vacancies). The electron mobility of the film was determined to be temperature independent at 0.01 cm2/Vs for the < 0001 > epitaxial film containing a Ti donor density of 4.0 x 1020 cm-3. Finally, the photoelectrochemical performance of these materials was tested by cyclic voltammetry and measurements of their quantum efficiencies. The 1% Ti doped hematite thin film exhibited the highest photocurrent density of these dense, thin films at 0.9mA/cm2 with an applied bias of 1.5V vs. RHE. The IPCE of this sample reached 15% at wavelengths between 300nm and 350nm after an annealing treatment at 580° for 36 h. The solid solution containing 33% ilmenite preformed nearly as well as the doped hematite. The performance decreased with higher ilmenite concentrations in the solid solution. For all samples containing any ilmenite, the onset potential shifted to lower values by ~200mV after the annealing treatment. The increase in charge carrier density upon reduction of Ti doped hematite was conformed by a Mott-Schottky analysis of the hematite/electrolyte interface. In contrast, only minor changes in the carrier density were observed when reducing an undoped hematite photoanode. Changes in slope of the Mott-Schottky plots revealed the presence of deep trap states in the hematite films. In-situ UV-vis spectroscopy displayed a pronounced optical signature corresponding to the existence of such deep levels. These results highlight the importance of carefully controlling photoanode processing conditions, even when operating within the material's extrinsic dopant regime, and more generally, provide a model for the electronic properties of semiconducting metal oxide photoanodes. |
Investigation of LbL assembly and M13 bacteriophage nanowires for DSSCs | A number of challenges related to the development of new organic-inorganic photovoltaic systems exist, including the ability to enhance the materials interface and improve the control required in development of nanoscale materials. Layer-by-layer (LbL) assembly allows for the incorporation of a wide range of functional materials into structured thin films based on the alternate adsorption of cationic and anionic species. Biomolecules, and in particular viruses, show great potential as components of functional materials due to their capacity for molecular recognition and self-assembly. Here we report that by substituting a negatively charged variant of M13 bacteriophage for the negatively charged polymer during the dip LbL assembly process, phage can be incorporated into a hybrid material with characteristics of both its biological and polymeric components. The resulting mesoporous polymer films can be used as a template for the construction of the titania photoanode of dye sensitized solar cells (DSSCs) with a novel nanowire architecture to enhance electron transport. The biotemplated nanowires are shown to significantly increase device electron diffusion length and increase device efficiency as compared to LbL-templated titania photoanodes made without bacteriophage. Spray LbL is also investigated as an assembly method for the construction porous templates for titania photoanodes. The necessary porous transition is shown to occur on flat substrates, like those normally utilized for DSSCs, and on porous metal meshes, substrates that have been proposed as lower-cost DSSC current collectors. Spray LbL is demonstrated to coat metal to different degrees of conformality as a function of mesh pore size. The conformality of the coating, in turn, determines which functions it could assume within a LbL-based DSSC. |
Spatial theory for biological design | Biological design is as ancient as human civilization. For thousands of years, living systems and natural processes have been manipulated by humans and their biological outcomes have been customized for different purposes. While the idea of biological design has always been prevalent throughout history, especially with the discovery of DNA, the ability to manipulate the form, function, and behavior of the living has significantly advanced. Today synthetic biology is pushing the frontiers of biological design even further. Now, living things can be completely abstracted from their original biological contexts, assembled like molecular constructs, and engineered like circuits or programmed like computational hardware. biological designers compose biological form and function by running modeling and simulation software; order standardized biological parts from online libraries and databases; utilize fabrication companies to synthesize gene products to prototype their designs; and build complex artifacts, applications and services that meet human wants, needs, fears, and desires on a daily basis. In this dissertation, I examine different practices of biological design in life sciences and engineering based on different theoretical models. I trace the history of information-based, relational, synthesis-oriented methods and present a new design framework that offers a spatial and a context-driven approach to the design of living matter. Being rooted in a different interpretation of space and spatiality in design, the framework approaches biological design systematically, at three stages: 1) the design of the basic units of the living (Units), 2) how different units are arranged and composed for different functionalities and behavior (Logic), and 3) the design of the biological contexts where biological artifacts live and perform their objectives (Context). This new framework intends to bring together a multitude of approaches from different design fields such as engineering, architecture and product Design that have their unique histories with living matter. The goal here is to demonstrate the ways different design paradigms can potentially shape our relationship with biological design in new ways; altering the design process, the objectives, the outcomes, and the social, cultural, and ethical perception of synthetic living. |
Design and analysis of fingernail sensors for measurement of fingertip touch fouce and finger posture | A new type of wearable sensor for detecting fingertip touch force and finger posture is presented. Unlike traditional electronic gloves, in which sensors are embedded along the finger and on the fingerpads, this new device does not constrict finger motion and allows the fingers to directly contact the environment without obstructing the human's natural haptic senses. The fingertip touch force and finger posture are detected by measuring changes in the coloration of the fingernail; hence, the sensor is mounted on the fingernail and does not interfere with bending or touching actions. Specifically, the fingernail is instrumented with miniature light emitting diodes (LEDs) and photodetectors in order to measure changes in the reflection intensity when the fingertip is pressed against a surface or when the finger is bent. The changes in intensity are then used to determine changes in the blood volume under the fingernail, a technique termed "reflectance photoplethysmography." By arranging the LEDs and photodetectors in a spatial array, the two-dimensional pattern of blood volume can be measured and used to predict the touch force and posture. This thesis first underscores the role of the fingernail sensor as a means of indirectly detecting fingertip touch force and finger posture by measuring the internal state of the finger. Desired functionality and principles of photoplethysmography are used to create a set of design goals and guidelines for such a sensor. |
Characterization of crater morphometry on the Moon and Mercury from altimetry observations | Recently acquired altimetry data from laser altimeters are used to assess the morphometry of impact craters. Data acquired by the Mercury Laser Altimeter on the MESSENGER spacecraft are used to measure the depths and diameters of 537 craters at the high northern latitudes on Mercury, including 42 polar-deposit-hosting craters (PDCs) which host material that is bright to earth-based radar observations. A comparative analysis suggests that the radar-bright material forms a thin (< 20 m) layer emplaced preferentially in comparatively young craters, contradicting an earlier morphometric study that indicated that PDCs contained a thick layer of water ice and dust. Topographic datasets from the lunar surface, collected by the Lunar Orbiter Laser Altimeter onboard the LRO spacecraft, are also used to evaluate the morphometry of 1,356 lunar craters. We study the morphologic change between the simple and complex crater regime, a manifestation of the transition between gravity-dominated and strength-dominated impact-forming processes, on the Moon and Mercury. The Moons transition diameter is near 16 km, in line with previous studies, while Mercurys is near 8 km, 2 km smaller than previously determined. The onset of gravity-dominated mechanisms scales inversely with gravity, which explains why Mercurys transition diameter is approximately half of the Moons. |
Wireless services business plan | The primary goals of this thesis were to analyze the market for wireless applications/services and create a business plan for a viable wireless application. There has been tremendous hype relating to wireless data services in the past few years with significant venture investment not yielding many commercially successful companies. In reviewing the wireless services market, there appears to be significant opportunity for the right wireless application/service to gain significant traction in the marketplace and spawn a successful startup. It was my assertion that one could create a service that allows users to find information from a community of users on a variety of technology platform that would prove to be the basis for a commercially viable startup. In writing the business plan for this startup, <ihufs>, I believe that there is a business justification for creating a company that will offer this service solving a specific consumer need. I hope to pursue this opportunity post graduation and further explore the creation of <ihufs> and the <ihufs> platform and service. |
Faceted Id/Entity : managing representation in a digital world | In this thesis, i articulate a theory of how and why individuals use context to convey only a facet of their identity in social interactions. Through this lens, i discuss current issues in digital identity management. In this discussion, i focus on the role of design in affecting an individual's ability to maintain control of personal representation and identity information. I argue that the architecture of current digital environments has altered our notions of context, motivating users to develop new mechanisms for managing their presentation. I take the stance that users should have the ability to control their digital identity for the same reasons that they seek to control their physical identity, most notably to present themselves in an appropriate manner in relation to the current situation. From this perspective, i argue for a design approach that will aid sociable designers in developing human-centered technologies that allow for individual control over personal identity. First, i argue the need for mechanisms of self-awareness and discuss what forms of awareness users should have. In doing so, i analyze current approaches to awareness and critique my own work on Social Network Fragments, a visualization tool for revealing the structure of one's digital social network. Alongside self-awareness, i present the need for identity management and critique my work on SecureId, a prototype intended to give users control over their digital presentation by offering security through identity-based knowledge. This thesis argues for empowering users through awareness and control, so that they may provide the level of regulation that is desirable. In doing so, i offer a novel approach to context and identity management in digital social interaction. |
Thermodynamics and kinetics of Mg intercalation for multivalent cathode applications | Energy storage, especially through electrochemical mechanisms such as batteries, is crucial for sustaining the ever-increasing energy needs of the future in a fossil-free manner. While the current industrial workhorse, lithium ion batteries, has shown tremendous improvements in energy and power-densities, via both materials selection and engineering advancements, the lithium ion technology is approaching the fundamental limits of what more can be achieved. Multi-valent (MV) chemistry, that pairs an energy-dense MV metal anode (such as Mg) with a high voltage cathode has the potential to surpass the energy densities achieved by current Li-ion batteries, along with improved safety and lower costs. However, moving into newer chemistries leads to newer challenges, such as developing cathodes that can reversibly intercalate Mg at high voltages, high rates and high capacities, apart from designing electrolytes that remain stable against both the electrodes. In this thesis, I focus on the challenge of MV cathode design and I explore the thermodynamic and kinetic properties of candidate oxide cathode materials for MV batteries, including polymorphs of V₂O₅, spinel-Mn₂O₄ and layered-Mg₂Mo₃O₈, using first-principles based methods. The undercurrent of the thesis is to obtain design principles that will aid in both optimization of existing cathodes and in the identification of new candidate materials. Utilizing a diverse set of tools, I benchmark the calculated properties, including average voltage curves, lattice parameters, cation-anion decorations in structures and activation barriers for Mg diffusion, to experimental observations, where possible. Finally, this thesis should serve as a guide for other computational-theorists and experimentalists, in the search for an energy-dense MV cathode that will in turn aid in the realization of a high energy density MV battery. |
Understanding social influence using network analysis and machine learning | If we are to enact better policy, fight crime and decrease poverty, we will need better computational models of how society works. In order to make computational social science a useful reality, we will need generative models of how social influence sprouts at the interpersonal level and how it leads to emergent social behavior. In this thesis, I take steps at understanding the predictors and conduits of social influence by analyzing real-life data, and I use the findings to create a high-accuracy prediction model of individuals' future behavior. The funf dataset which comprises detailed high-frequency data gathered from 25 mobile phone-based signals from 130 people over a period of 15 months, will be used to test the hypothesis that people who interact more with each other have a greater ability to influence each other. Various metrics of interaction will be investigated such as self-reported friendships, call and SMS logs and Bluetooth co-location signals. The Burt Network Constraint of each pair of participants is calculated as a measure of not only the direct interaction between two participants but also the indirect friendships through intermediate neighbors that form closed triads with both the participants being assessed. To measure influence, the results of the live funf intervention will be used where behavior change of each participant to be more physically active was rewarded, with the reward being calculated live. There were three variants of the reward structure: one where each participant was rewarded for her own behavior change without seeing that of anybody else (the control), one where each participant was paired up with two 'buddies' whose behavior change she could see live but she was still rewarded based on her own behavior, and one where each participant who was paired with two others was paid based on their behavior change that she could see live. As a metric for social influence, it will be considered how the change in slope and average physical activity levels of one person follows the change in slope and average physical activity levels of the buddy who saw her data and/or was rewarded based on her performance. Finally, a linear regression model that uses the various types of direction and indirect network interactions will be created to predict the behavior change of one participant based on her closeness with her buddy. In addition to explaining and demonstrating the causes of social influence with unprecedented detail using network analysis and machine learning, I will discuss the larger topic of using such a technology-driven approach to changing behavior instead of the traditional policy-driven approach. The advantages of the technology-driven approach will be highlighted and the potential political-economic pitfalls of implementing such a novel approach will also be addressed. Since technology-driven approaches to changing individual behavior can have serious negative consequences for democracy and the free-market, I will introduce a novel dimension to the discussion of how to protect individuals from the state and from powerful private organizations. Hence, I will describe how transparency policies and civic engagement technologies can further this goal of 'watching the watchers'. |
Tools for S-C0₂ cycle analysis and the cycle's applicability to SFRs | The Sodium-Cooled Fast Reactor (SFR) and the Supercritical Carbon Dioxide (S-C0₂) Recompression cycle are two technologies that have the potential to impact the power generation landscape of the future. In order for their implementation to be successful, they must compete economically with existing light water reactors and the conventional Rankine cycle. Improvements in efficiency, while maintaining safety and proliferation goals, will allow the SFR to better compete in the electricity generation market. These improvements will depend on core design as well as the balance of plant, including the choice of steam or C0₂ as the working fluid. This work has developed some of the tools necessary for evaluating different design core and balance of plant options. Much of it has concentrated on the S-C0₂ Recompression cycle. S-C0₂ promises to be useful as a working fluid in high-efficiency power conversion systems for SFRs because it achieves higher efficiencies at the high temperatures associated with SFRs. The recompression cycle is capable of operating with very high efficiencies due to the low compressor work needed when C0₂ approaches its critical point at the compressor inlet. The potential of this cycle to meet the needs of next-generation plants must be investigated across the entire range of operations and within each component of the system. A steady-state code for analysis of the recompression cycle was previously developed at MIT in the form of CYCLES II, but the present work has made significant improvements to this code that make the new version, CYCLES III, more versatile. |
Development, implementation and analysis of the first recycling process for alkaline liquid metal batteries | Increasing energy prices, new environmental laws and geopolitical interests demand for new, more efficient and cheaper grid level energy storage solutions. Grid level energy storage refers to large scale energy storage applications that are connected to the power grid. Ambri Inc. is a MIT startup that develops liquid metal batteries for grid level energy storage. Their liquid metal battery operates at elevated temperatures and uses molten metals as electrodes thereby exhibiting a very low fade rate over hundreds of charging and discharging cycles. Ambri cooperated with MIT to develop a new recycling process for their unique battery chemistry to implement a sustainable end of life management for their product. This thesis describes the process development, implementation and analysis of a hydrometallurgical recycling process for a liquid metal battery. According to jointly developed process requirements, the MIT team build a process that is capable of recycling 5 liquid metal batteries per batch with an estimated processing time of 60 minutes. This will increase Ambri's profit by several hundred thousands of dollars even during the first year of production. The performed analysis of the process investigated safe and stable operating conditions, cost efficiency and scalability. The MIT team concluded that the newly developed recycling process best accommodates for Ambri's current needs and future growth compared to the only competing process, the full cell incineration with following hazardous waste landfill deposition. |
Microbial community structure and dynamics on patchy landscapes | Microbes are tiny metabolic engines with large-scale effects on industry, the environment, and human health. Understanding how the micron-scale actions (and interactions) of individual microbes give rise to macro-scale consequences remains a major challenge in microbial ecology. However, for the most part, studies employ coarsegrained sampling schemes, which average over the heterogeneous microscopic structure of microbial communities. This has limited our ability to establish mechanistic links between dynamics occurring across these disparate spatial scales. However, such links are critical for (a) making sense of the tremendous extant microbial diversity on Earth, and (b) predicting how perturbations (e.g., global climate change) may influence microbial diversity and function. In this thesis, I characterize the structure and dynamics of wild bacterial populations in the ocean at spatial scales of tens of microns. I then employ a simple, two-strain laboratory model system to link (cooperative) inter-species interactions at local scales to emergent properties at larger scales, focusing on spatially connected meta-communities undergoing range expansions into new territory. This work encompasses diverse environments (ranging from well-mixed communities in the laboratory to individual crustaceans) and approaches (including mathematical modeling, highthroughput sequencing, and traditional microbiological experiments). Altogether, we find that the microscale environment inhabited by a microbe - that is, "what the neighborhood is like" and "who lives next to whom" - shapes the structure and dynamics of wild microbial populations at local scales. Moreover, these local interactions can drive patterns of biodiversity and function, even at spatial scales much larger than the length of an individual cell. Thus, our work represents a small step toward developing mechanistic theories for how microbes shape our planet's ecosystems. |
Matching and compressing sequences of visual hulls | In this thesis, we implement the polyhedral visual hull (PVH) algorithm in a modular software system to reconstruct 3D meshes from 2D images and camera poses. We also introduce the new idea of visual hull graphs. For data, using an eight camera synchronous system after multi-camera calibration, we collect video sequences to study the pose and motion of people. For efficiency in VH processing, we compress 2D input contours to reduce te number of triangles in the output mesh and demonstrate how subdivision surfaces smoothly approximate the irregular output mesh in 3D. After generating sequences of visual hulls from source video, to define a visual hull graph, we use a simple distance metric for pose by calculating Chamfer distances between 2D shape contours. At each frame of our graph, we store a view independent 3D pose and calculate the transition probability to any other frame based on similarity of pose. To test our approach, we synthesize new realistic motion by walking through cycles in the graph. Our results are new videos of arbitrary length and viewing direction based on a sample source video. |
Stakeholder value network analysis for the mobile services ecosystem | The mobile services ecosystem has evolved and continues to evolve at a rapid pace adjusting to the different players competing to be part of the value creation and capture. This thesis attempts to capture a holistic view of the entire ecosystem and performs a stakeholder value network analysis on the ecosystem and its members. The qualitative model lists, maps and identifies the stakeholders and the monetary, goods / services, information, participation and public benefit value flows between the stakeholders in the ecosystem. The quantitative model attaches numerical values to the different value flows based on a characterization of needs and the importance of the sources fulfilling those needs and creates a value network that is then analyzed to produce relative importance of stakeholders from various perspectives and a ranking of the importance of different types of direct and indirect value flow loops in the system thus creating a framework for strategic ecosystem analysis. |
Design of a model pipeline for testing of piezoelectric micro power generator for the Trans-Alaska Pipeline System | In order to provide a reliable corrosion detection system for the Trans-Alaska Pipeline System (TAPS), a distributed wireless self-powered sensor array is needed to monitor the entire length of the pipeline at all times. Such a sensor faces two primary challenges: a method to provide power for the sensor, and a method to detect corrosion. This project has two goals: to build a model of the TAPS as a test bed for a piezoelectric micro power generator (PMPG), and to use the model to explore corrosion detection methods (perhaps by analyzing changes in the vibration spectrum), for use in the sensor array. To miniaturize the TAPS while maintaining its vibration spectrum, we will specify the dimensions of the model to have the same natural frequency, turbulent flow, and vortex induced vibrations as the actual pipeline. The model will serve as a test bed for various PMPG designs, and also serve as a starting point for exploring methods to detect corrosion in pipes. The primary vibration mode was found to be due to the natural frequency of the pipe, which was 20.2 Hz for the TAPS. Experimentally, we found the frequency to be in a range from 12-19 Hz. PMPG devices for use in the TAPS should be tuned to this frequency range. |
Growth and characterization of bismuth perovskite thin films for integrated magneto-optical isolator applications | In this thesis, we discuss the motivation for integrated magneto-optical isolators and explain why the orthoferrite is such an attractive materials class for this purpose. We then derive from first physical principles the dependence of Faraday rotation, absorption, and certain figures of merit on the material's dielectric tensor elements. Next, we use pulsed laser deposition to grow thin films of BiFeO3 on MgO (001) and SrTiO3 (001) substrates. After optimizing growth conditions to obtain high quality films, we characterize the films' crystal structure with two-dimensional x-ray diffraction. We then examine the magnetic, optical, and magneto-optical properties of these films. We find that the highly textured films grown on SrTiO3 are monoclinic with an out-of-plane c-axis aligned with the (001) direction of the substrate and approximate pseudocubic lattice parameters of a = b = 4.04 A, c = 3.95 A, and 90° - [beta] = -0.88°. These films are weakly magnetic, with a magnetization of 1.2 emu/cm3 at an applied field of 10 kOe; highly absorptive, with an average absorption coefficient of 910 cm-1; and possess a low specific Faraday rotation of 320/cm at 1.8 kOe of applied field. As expected, we find that the magneto-optical figure of merit is negligible for this material due to its high absorption, which we attribute to a thin surface layer of phase separated bismuth and iron oxides caused by bismuth segregation during growth. We offer additional explanations for these values and show the first results of newer, more promising work with mixed cation perovskite. |
Application of overset grids for aerodynamic assessment of an advanced civil transport aircraft | Simulations are presented for 1:20 and 1:11 scale configurations of an advanced civil transport designed to use boundary layer ingestion (BLI). Comparison with wind tunnel results on unpowered configurations show that the computed and measured lift slopes agree within 15% and the drag at simulated cruise point to within 20%. Computations have been carried out for an initial and a redesigned configuration of the aft section of the aircraft. The redesigned version showed a reduction in lift and an increase in drag compared to the initial design. The mechanical power for a 1:11 powered configuration at cruise, without BLI, has also been calculated. |
Turbulent fluid jet excavation in cohesive soil : with particular application to jet grouting | This thesis reviews the jet grouting methodology, and the current state of practice and research. Current methods of prediction of jet grout diameters are highly empirical and site specific, and do not take into account the jet hydrodynamics and soil properties explicitly. A rational model to describe the jet excavation mechanism for cohesive soil is presented in this thesis, with the aim of providing an improved prediction tool that can be used in jetting design. The model is based on the assumption that the velocity distribution in the cutting jet is equivalent to that of a free jet with boundaries corresponding to the dimensions of the cavity excavated in the ground. The shape of the cavity formed depends on the erosional properties of the soil and jet expansion is limited by the resistance at the jet-soil interface. It is hypothesized that the shape of the cavity excavated follows the locus of the jet radius satisfying the condition of constant wall shear stress. The model predicts that the limit of jet penetration is reached when the dynamic pressure at the jet tip becomes equal to the ultimate soil bearing resistance. The model was validated by laboratory jetting tests, using soil specimens manufactured from powdered kaolin clay, cement and water. Cement-soil ratios (CSR) of 2.5 to 7.5% were used to produce specimens with undrained shear strength (Su) ranging from 5 to 45 kPa. A period of at least 3 days was allowed for the specimen to cure in the test tank before jetting was commenced. The tests were conducted using different nozzle diameters, jetting pressures, rotation speeds and soil strengths. The specimens were exhumed after completion of each test to map the shape of the cuts excavated. |
A spectral approach to noninvasive model-based estimation of intracranial pressure | Intracranial pressure (ICP) is the hydrostatic pressure of the cerebrospinal fluid. Ideally, ICP should be monitored in many neuropathological conditions, as elevated ICP is correlated with poor neurocognitive outcomes after injuries to the brain. Measuring ICP requires the surgical placement of a sensor or catheter into the brain tissue or cerebrospinal fluid spaces of the brain. With the risk of infection and brain damage, ICP is only measured in a small subset of those patients whose treatment could benefit from knowing ICP. We expand on a previously proposed model-based time-domain approach to noninvasive, patient-specific and continuous estimation of ICP using routinely measured waveforms that has been validated on patients with traumatic brain injuries. Here, we present a model-based algorithm to estimate ICP using the functional relationship between the spectral densities of the routinely measured waveforms. We applied this algorithm to both a simulated and clinical dataset. For the simulated dataset, we achieved a mean error (bias) of 1.2 mmHg and a standard deviation of error (SDE) of 2.2 mmHg. For the clinical dataset of patients with traumatic brain injuries, we achieved a bias of 13.7 mmHg and a SDE of 15.0 mmHg. While the clinical results are not favorable, we describe sources of estimation error and future directions of research to improve the ICP estimates. |
Catalysis and manufacturing of two-scale hierarchical nano- and microfiber advanced aerospace fiber-reinforced plastic composites | The development of hierarchical nanoengineered "fuzzy fiber" aerospace fiber-reinforced plastic (FRP) composite laminates holds the potential for enabling future generations of lightweight, durable, and multifunctional vehicle structures. By reinforcing the weak matrix-rich regions between individual fibers and plies, the circumferential growth of aligned carbon nanotubes (A-CNTs) on carbon microfibers (CFs) enables new composites with improved strength, toughness, electrical and thermal properties. While these improvements have been empirically demonstrated on alumina fiber FRPs, CNT growth degrades the CFs and sacrifices in-plane FRP properties for the benefits of CNT reinforcement. This thesis presents novel and scalable methods for realizing advanced fuzzy carbon fiber reinforced plastic (fuzzy CFRP) composite laminates with retained CF and interlaminar strength properties. Earth-abundant sodium (Na) is revealed as a new facile catalyst for CNT growth that allows for direct deposition of the catalyst precursor on carbon fabrics without any fiber pretreatments. This new catalyst discovery also enables high-yield CNT growth on a variety of low-temperature substrates. Simultaneously, this finding has led to other novel findings in carbon nanostructure catalysis including a core-shell morphology and the use of other alkali metals (e.g., potassium) for CNT growth. Towards the development of advanced composites, vacuum-assisted resin infusion processes are studied and refined, resulting in high-quality woven and unidirectional fuzzy (via Na-catalysis of CNTs) CFRP laminates. Growth uniformity improvement studies yielded strategies for increasing the quantity of CNT reinforcement within matrix-rich regions. Moreover, a new commercial unidirectional fabric enables the first retention of CF properties concomitant with interlaminar shear strength retention in the fuzzy CFRP architecture. The contributions of this thesis extend beyond CF composites: techniques developed for improving fuzzy CF synthesis were applied towards demonstrating A-CNT growth on SiC woven fabric, desired for creating damage tolerant and multifunctional lightweight vehicle systems. These advances pave the way for improvements in catalysis of nanostructures, electronics interfaces, energy storage devices, and advanced composite materials. |
How urban design can promote care for ecological systems | Human behavior is, in aggregate present state, unsustainable by the ecosystem of the earth. This thesis develops a behavior change model of how societies can move from unsustainable to sustainable behavior, and, because behavior is in large part a result of personal conviction, the factors that result in voluntary behavior change towards more environmentally sustainable behavior. The three key factors that trigger a desire to behave sustainably are intellectual understanding of ecosystems, engagement with ecosystems, and feedback on personal ecosystem impacts. Urban design is a key element in how the ecosystem is presented and culturally interpreted, and this thesis examines how urban design can be used to promote care for the ecosystem by offering experiences that provide the factors of behavior change. Several example cases are provided, as well as an examination of how we sense natural systems and design suggestions. |
System for identifying and reporting changes to course websites | CourseDiff is a prototype system that periodically samples course websites and notifies users via email when it identifies changes to those sites. The system was developed after conducting a study of 120 web pages from 50 MIT course websites sampled for two months during the spring semester of 2009. The study found that only 18% of changes to the HTML content of course website data are actually important to the content of the page. A closer examination of the corpus identified two major sources of trivial changes. The first is automatically generated content that changes on every visit to the page. The second is formatting and whitespace changes that do not affect the page's textual content. Together, these two sources produce over 99% of the trivial changes. CourseDiff implements an algorithm to filter out these trivial changes from the webpages it samples and a change reporting format for the changes that are identified as important. A small user test on part of the CourseDiff interface indicated that the system could feasibly be used by students to track changes to course websites. |
Enhancement of fine particle deposition to permeable sediments | Predictions of deposition rate are integral to the transport of many constituents including contaminants, organic matter, and larvae. Review of the literature demonstrates a general appreciation for the potential control of deposition by bed roughness, but no direct tests involving flat sediment beds. Understanding the mechanisms at work for flat sediment beds would provide the basis for exploring more complicated bed conditions and the incorporation of other transport processes, such as bioturbation and bedload transport. Generally, fine particle deposition rates are assumed to be equivalent to the suspension settling velocity, therefore, deposition rates in excess of settling are considered enhanced. Flume observations of deposition were made using treatments that covered a wide range of flow, particle, and bed conditions. Specific treatments demonstrated large enhancements (up to eight times settling). Delivery of particles to the interface is important, but models based on delivery alone failed to predict the observed enhancement. This necessitated the development of a new model based on a balance between delivery and filtration in the bed. Interfacial diffusion was chosen as a model for particle delivery. Filtration of particles by the bed is a useful framework for retention, but the shear in the interstitial flow may introduce additional factors not included in traditional filtration experiments. |
Media informed architecture | In todays society we are in a continuous state of distraction. Our cell phones and MP3 players provide us with a steady stream of information and imagery to deter our mind fro the disruption of advertisements and solicitations surrounding us.This thesis explores the relationship between media and architecture. Specifically, how the moving image and the experience of moving through built space can direct, distract and alter perception. Interaction between the public and the displayed media create an environment that is both social and engaging. this relationship is considered within a mobile media park sited inside the Los Angles river channel in downtown LA. |
Stochastics : a Bayesian network architecture for combined user modeling, sensor fusion, and computational storytelling for interactive spaces | This thesis presents a mathematical framework for real-time sensor-driven stochastic modeling of story and user-story interaction, which I call sto(ry)chastics. Almost all sensor-driven interactive entertainment, art, and architecture installations today rely on one-to-one mappings between content and participant's actions to tell a story. These mappings chain small subsets of scripted content, and do not attempt to understand the public's intention or desires during interaction, and therefore are rigid, ad hoc, prone to error, and lack depth in communication of meaning and expressive power. Sto(ry)chastics uses graphical probabilistic modeling of story fragments and participant input, gathered from sensors, to tell a story to the user, as a function of people's estimated intentions and desires during interaction. Using a Bayesian network approach for combined modeling of users, sensors, and story, sto(ry)chastics, as opposed to traditional systems based on one- to-one mappings, is flexible, reconfigurable, adaptive, context-sensitive, robust, accessible, and able to explain its choices. To illustrate sto(ry)chastics, this thesis describes the museum wearable, which orchestrates an audiovisual narration as a function of the visitor's interests and physical path in the museum. The museum wearable is a lightweight and small computer that people carry inside a shoulder pack. It offers an audiovisual augmentation of the surrounding environment using a small eye-piece display attached to conventional headphones. The wearable prototype described in this document relies on a custom-designed |
Heparan sulfate glycosaminoglycan regulation of vasculogenesis | Neovascularization is an essential process to repair ischemic tissues following myocardial infarction, stroke, diabetic complications, or transplant procedures. Blood vessels are generated by distinct vasculogenic and angiogenic processes. Although multiple proangiogenic factors have been identified, limited success has been achieved translating these as clinical therapeutics. Furthermore, recent studies have shown that vasculogenesis contributes to adult neovascularization in multiple settings. Harnessing the vasculogenic potential of embryonic stem cells is an emerging concept to generate neovasculature. The differentiation of embryonic stem cells into endothelium has been well documented, however most studies focus on genetic or chemokine regulation. Limited information exists which implicates the role of the extracellular microenvironment in stem cell differentiation. Heparan sulfate glycosaminoglycans (HSGAG) are a crucial part of the dynamic extracellular matrix and have been shown to regulate multiple signaling cascades, including vasculogenic specific growth factors VEGF and FGF. The goal of this thesis is to elucidate the role of HSGAG in vasculogenesis. An embryonic stem cell embryoid body model was used to establish the necessity of sulfated HSGAG for endothelial differentiation. We identified that the chemical composition of HSGAG sulfation patterns change with differentiation. Perturbation of HSGAG structure by chemical, enzymatic, or genetic modification effectively inhibited vasculogenesis. Genetic silencing of HSGAG modifying enzyme, N-deacetylase/N-sulfotransferase-1, translated to inhibition of HSGAG sulfation and resulted in impaired blood vessel development in zebrafish embryos. Interestingly, vessel formation in both embryonic stem cell and zebrafish models was restored by the addition of exogenous HSGAG, opening the door for engineering glyco-based microenvironments for controlling vascular development. To explore novel mechanisms of vasculogenesis modulated by HSGAG perturbation, we performed a global transcriptome analysis of N-deacetylase/N-sulfotransferase-1 mutant zebrafish embryos. Several novel pathways were identified that regulate vascular differentiation, including Foxo3A and Insulin-Like Growth Factor (IGF) pathways. We explored the role of IGFs in vasculogenesis specifically and determined for the first time that IGF1 and IGF2 promote mesoderm and endothelial differentiation, mediated through HIFl[alpha] stabilization, in embryonic stem cells. In summary, we've identified several mechanisms by which HSGAG regulate neovascularization, laying the groundwork for incorporating HSGAG in strategies for ischemic tissue regeneration. |
The real estate user charge and its impact on housing development in China | Off-budget public finance in China is an informal system of levies that emerges in response to local fiscal needs. Off-budget funds serve as part of local government revenue and help finance local government expenditures. In the real estate sector, local governments raise off-budget funds by charging many types of user fees on housing development. The user charge revenue is an important revenue source to finance urban public service expenditures. Some drawbacks of user charges such as legal ambiguity and the lack of transparency reflect noteworthy investment risks in Chinese housing development. Thus, a good understanding of the real estate user charges is critical for foreign and domestic developers who are interested in housing development in China. In this study, we have explored literatures about urban public service policies, presented facts about the current user charge practices, analyzed the costs incurred in use fee transactions, evaluated the efficiency and equity contributions of the current system, and illustrated important issues in details by a case study of a single user fee. |
Racial, not rational : economic threat, symbolic racism, and affirmative action | For decades, scholars have debated the determinants of whites' attitudes about racialized policies such as welfare, busing, and affirmative action. While some have argued that whites formulate their positions rationally according to perceived economic threat, others have asserted that such policy attitudes are the function of one's level of symbolic racism, with little to no influence from economic considerations. Using data from the 2012 Cooperative Congressional Elections Study and demographic data, I assess the effects of actual economic competition and an individual's other attitudes on white opposition to affirmative action. Furthermore, in order to identify the levels, if any, through which the economic threat mechanism operates, this paper measures economic threat in several different ways: at both the level of the individual and the level of whites as a group, and each of these at both the zip code and county levels. I find strong support for the symbolic racism theory of policy attitude formation, as respondent attitudes are driven mostly by racial affect, ideology, and party identification. No matter the level at which economic threat is measured, objective economic conditions do not seem to influence one's attitudes about affirmative action. |
A market and engineering study of a 3-kilowatt class gas turbine generator | Market and engineering studies were performed for the world's only commercially available 3 kW class gas turbine generator, the IHI Aerospace Dynajet. The objectives of the market study were to determine the competitive requirements for small generators in various U.S. applications, assess the unit's current suitability for these applications, and recommend ways to modify performance or marketing practices to make it more competitive. Engineering study goals included developing an accurate cycle model and assessing the potential for performance improvement. The market study found that the current high selling price precludes competitiveness in most segments of the U.S. civil market. One potential exception may be the marine market, where price sensitivity is less of an issue and a premium is paid for quiet operation, a distinct advantage of the Dynajet. A gas turbine generator solution has more potential in the military market, where the difference from incumbent prices is smaller than in the civil market. The Dynajet is also an appealing military solution because of its high reliability and quiet operation. The market study concluded that increasing power output and efficiency while reducing purchase price would be the most effective approach to improved competitiveness. Alternatively, the current strengths could be leveraged by adapting it for use with an absorption cooler and by emphasizing its superior emission characteristics to consumers and regulators. The engineering study discovered that cycle performance is degraded by secondary nonidealities including flow leakage, heat leakage, and thermal flow distortion. Although these nonidealities are present to some degree in all gas turbines, their impacts are larger in small-scale engines. |
Electromagnetic energy harvester and self-powered embedded system | Energy harvesting offers an important design option for creating sensing and control elements without a requirement for custom wiring or batteries. The independent and care-free nature of energy harvesting enables monitoring devices to penetrate wider and deeper into our daily lives, making accommodation of fine sensing and control for condition-based maintenance ever more feasible. Finer granularity in sensing and control, which is the future of energy efficiency, alone is an immense benefit as it can reduce time and cost associated with a potential repair. Combined with condition-based maintenance, it can prevent potential down-time of a machine under monitoring. An exciting possibility creates a "self-powered" embedded system with an integrated energy harvester for electromechanical diagnosis. This non-intrusive energy harvester is designed to extract energy from magnetic fields around a power line of a load, in the manner of a current transformer. In contrast to the conventional usage of magnetic elements, such as transformers and inductors, the analysis on this "current transformer" reveals a critical result: for any given core for any particular application, power harvest is maximized when the core is permitted to saturate at an opportune time in the line cycle. The design of this integrated energy harvester is fully explored in the thesis, including: development of new models to incorporate a fully saturating magnetic core for simulation; designs of power electronics circuits for maximizing power harvest; and integration of the harvester into the embedded system as a practical power supply. The design of a self-powered and low-power embedded system, vibration assessment monitoring point with integrated recovery of energy (VAMPIRE), is discussed in depth in the thesis. The overall architecture of the embedded system is first presented, followed by designs of individual subsystems, the power package and the sensor package. In the power package, initialization, energy buffer, power interfaces, power regulation, and microcontroller design are explored. In the sensor package, power budget, sensors, data storages, storage management, wireless communication, and corresponding user interfaces are explored. Finally, impedance spectroscopy for an electromechanical load is discussed. Using the electrical and vibrational data that are nonintrusively collected from electromagnetically self-powered embedded system, structural issues of the load, i.e., changes in the stiffness of mounts and the imbalance of a shaft, can be clearly identified, making it feasible for this self-powered embedded system to be used for condition-based maintenance. |
Machine learning approaches to challenging problems : interpretable imbalanced classification, interpretable density estimation, and causal inference | In this thesis, I address three challenging machine-learning problems. The first problem that we address is the imbalanced data problem. We propose two algorithms to handle highly imbalanced classification problems. The first algorithm uses mixed integer programming to optimize a weighted balance between positive and negative class accuracies. The second method uses an approximation in order to assist with scalability. Specifically, it follows a characterize-then-discriminate approach. The positive class is first characterized by boxes, and then each box boundary becomes a separate discriminative classifier. This method is computationally advantageous because it can be easily parallelized, and considers only the relevant regions of the feature space. The second problem is a density estimation problem for categorical data sets. We present tree- and list- structured density estimation methods for binary/categorical data. We present three generative models, where the first one allows the user to specify the number of desired leaves in the tree within a Bayesian prior. The second model allows the user to specify the desired number of branches within the prior. The third model returns lists (rather than trees) and allows the user to specify the desired number of rules and the length of rules within the prior. Finally, we present a new machine learning approach to estimate personalized treatment effects in the classical potential outcomes framework with binary outcomes. Strictly, both treatment and control outcomes must be measured for each unit in order to perform supervised learning. However, in practice, only one outcome can be observed per unit. To overcome the problem that both treatment and control outcomes for the same unit are required for supervised learning, we propose surrogate loss functions that incorporate both treatment and control data. The new surrogates yield tighter bounds than the sum of the losses for the treatment and control groups. A specific choice of loss function, namely a type of hinge loss, yields a minimax support vector machine formulation. The resulting optimization problem requires the solution to only a single convex optimization problem, incorporating both treatment and control units, and it enables the kernel trick to be used to handle nonlinear (also non-parametric) estimation. |
Software tools for the graphic designer | This thesis describes three packages of programs of interest to the graphic designers: (l) a tool for pointing at objects on the display monitor, (2) a tool for organizing sets of programs into interactive menus, and (3) a tool for digitizing type fonts. The graphic designer at the computer uses at least three spaces at once: the page he is designing, the device space of the display monitor, and the device space of the graphics tablet. The first tool figures the transforms necessary to map one space into another. This allows the user to point at a location on the tablet and thereby identify corresponding points on the screen and in user-defined space. Interactive page layout requires the orchestration of large numbers of separate procedures. Choosing between these procedures can be made easier for the user if he is presented with a menu of options when a decision is required. The second tool is a package of programs to help the programmer create and manage simple or complex menus. Simulating type on a monitor requires digitized fonts. The third tool is a package of programs which extracts the edge of a letter-form image in the frame buffer and stores it as a chain code. |
Addressing deep uncertainty in space system development through model-based adaptive design | When developing a space system, many properties of the design space are initially unknown and are discovered during the development process. Therefore, the problem exhibits deep uncertainty. Deep uncertainty refers to the condition where the full range of outcomes of a decision is not knowable. A key strategy to mitigate deep uncertainty is to update decisions when new information is learned. NASA's current uncertainty management processes do not emphasize revisiting decisions and therefore are vulnerable to deep uncertainty. Examples from the development of the James Webb Space Telescope are provided to illustrate these vulnerabilities. In this research, the spacecraft development problem is modeled as a dynamic, chance-constrained, stochastic optimization problem. The Model-based Adaptive Design under Uncertainty (MADU) framework is introduced, in which conflict-directed search is combined with reuse of conflicts to solve the problem efficiently. |
Protocols for multi-antenna ad-hoc wireless networking in interference environments | A fundamental question for the design of future wireless networks concerns the nature of spectrum management and the protocols that govern use of the spectrum. In the oligopoly model, spectrum is owned and centrally managed, and the protocols tend to reflect this centralized nature. In the common's model, spectrum is a public good, and protocols must support ad hoc communication. This work presents the design, tradeoffs and parameter optimization for a new protocol (Simultaneous Transmissions in Interference (STI-MAC)) for ad hoc wireless networks. The key idea behind the STI-MAC protocol is 'channel stuffing,' that is, allowing network nodes to more efficiently use spatial, time and frequency degrees of freedom. This is achieved in three key ways. First, 'channel stuffing' is achieved through multiple antennas that are used at the receiver to mitigate interference using Minimum-Mean-Squared-Error (MMSE) receivers, allowing network nodes to transmit simultaneously in interference limited environments. The protocol also supports the use of multiple transmit antennas to beamform to the target receiver. Secondly, 'channel stuffing' is achieved through the use of a control channel that is orthogonal in time to the data channel, where nodes contend in order to participate on the data channel. And thirdly, 'channel stuffing' is achieved through a protest scheme that prevents data channel overloading. The STI-MAC protocol is analyzed via Monte-Carlo simulations in which transmitter nodes are uniformly distributed in a plane, each at a fixed distance from their target receiver; and as a function of network parameters including the number of transmit and receive antennas, the distance between a transmitter-receiver pair (link-length), the average number of transmitters whose received signal is stronger at a given receiver than its target transmitter (link-rank), number of transmitter-receiver pairs, the distribution on the requested rate, the offered load, and the transmit scheme. The STI-MAC protocol is benchmarked relative to simulations of the 802.11(n) (Wi-Fi) protocol. The key results of this work show a 3X gain in throughput relative to 802.11(n) in typical multi-antenna wireless networks that have 20 transmitter-receiver pairs, a link-length of 10 meters, four receive antennas and a single transmit antenna. We also show a reduction in delay by a factor of two when the networks are heavily loaded. We find that the link-rank is a key parameter affecting STIMAC gains over Wi-Fi. In simulations of networks with 40 transmit-receiver pairs, link-rank of three, a link-length of 10 meters, and eight transmit and receive antennas in which the transmitter beamforms to its target receiver in its strongest target channel mode, we find gains in throughput of at least 5X over the 802.11(n) protocol. |
Unintended effects of federal transportation policy: a look at the lifecycle costs of the interstate system | The federal surface transportation program in the United States stands at a crossroads. Short on money and lacking vision, the program is in need of reevaluation and reform. This thesis attempts to illustrate the current issues affecting the program within the historical context of transportation system development and proposes a series of policy reforms to refocus the program. An analysis of the history of US transportation policy demonstrates ongoing persistent trends towards multimodal approaches, increased federal involvement and a shift in focus from rural development to enabling the growth of metropolitan economies. The analysis also shows how Congress has historically made significant progress when an over-arching vision can be connected to implementation mechanisms that provide new funding to broad constituencies in all fifty states. By positing a basic model utilizing infrastructure lifecycle costs to illustrate the dynamics of systematic infrastructure needs, the research demonstrates the added costs of past policies - particularly deferred maintenance - and the implications of current inaction. The research finds an estimated $7.2 billion dollar shortfall in annual funding on the Interstate system alone - mostly for reconstruction. Additionally, the results indicate a significant increase in annual system costs ($28.3 billion vs. $18.82 billion) under a regime of deferred maintenance as opposed to regular upkeep. Using this as a foundation, the balance of the work discusses the political argument in support of a federal role in system maintenance, recommends a series of policy reforms to address short and long term issues with the federal program and presents an overview of possible revenue streams to fund these changes. The thesis then recommends that a new national multimodal vision, focused on state of good repair and the opportunity to generate short-term construction jobs and long-term economic growth, can be the basis for successful reauthorization legislation. |
Using demand signal repository to forecast demand | Improving forecast accuracy has positive effects on supply chain performance. Forecast accuracy can reduce inventory levels, increase customer service levels and responsiveness, or a combination of the two. However, the further upstream in the supply chain, the more difficult it becomes to forecast accurately. Demand for consumer products might be subject to factors that are hard to identify and quantify. One way to overcome this is to observe external factors or predictors that might help explain demand. The purpose of this thesis is to explore the factors that potentially influence the demand of a fast-moving consumer product (bottled water), and build a demand signal repository for these factors to help the manufacturer generate more accurate forecasts. We identified more than 30 such factors that might affect demand, using interviews and industry research. We tested more than 200 causal models of the relationship between observed demand and the predicting factors. The resulting model explained almost 60% of demand for two out of three customers using daily buckets and over 85% using weekly buckets compared to less than 50% using time-series techniques. Using the results of this extensive analysis, we propose a new forecasting model. We also identified additional factors that could not be included this analysis due to the lack of data; adding these to the model may further improve the forecast accuracy. |
Column generation approaches to the military airlift scheduling problem | In this thesis, we develop methods to address airlift scheduling, and in particular the problem of scheduling military aircraft capacity to meet ad hoc demand. Network optimization methods typically applied to scheduling problems do not sufficiently capture all necessary characteristics of this problem. Thus, we develop a new method that uses integer linear programming (IP) with column generation to make the problem more tractable while incorporating the relevant characteristics. In our method, we decompose the problem into two steps: generating feasible aircraft routes, and solving the optimization model. By ensuring that routes are feasible with respect to travel time, ground time, crew rest, and requirement restrictions when we build them, we do not need to encode these characteristics within the IP optimization model, thus reducing the number of constraints. Further, we reduce the number of decision variables by generating only the fraction of feasible aircraft routes needed to find near-optimal solutions. We propose two methods for generating routes to include in the IP model: explicit column generation and selective column generation. In explicit column generation, all aircraft routes that we could potentially consider including in the model are generated first. Starting with a subset of these routes, we iteratively use reduced cost information obtained by solving a relaxed version of the IP model to choose more routes to add from the original set of routes. In selective column generation, we first generate a small set of feasible aircraft routes. Starting with this set of routes, we iteratively generate more routes by solving a relaxed version of the IP model and then combine routes in the solution together and add those that are feasible to the route set. In both methods, we iterate until there are either no other routes to include or the solution stops improving. Last, we solve the IP model with the final set of routes to obtain an integer solution. We test the two approaches by varying the number of locations in the network, the number of locations that are wings, and the number of requirements. We show that selective column generation produces a solution with an objective value similar to that of explicit column generation in a fraction of the time. In our experiments, we solve problems with up to 100 requirements using selective column generation. In addition, we test the impact of integrating lines of business while scheduling airlift and show a significant improvement over the current process. |
Computational imaging with small numbers of photons | The ability of an active imaging system to accurately reconstruct scene properties in low light-level conditions has wide-ranging applications, spanning biological imaging of delicate samples to long-range remote sensing. Conventionally, even with timeresolved detectors that are sensitive to individual photons, obtaining accurate images requires hundreds of photon detections at each pixel to mitigate the shot noise inherent in photon-counting optical sensors. In this thesis, we develop computational imaging frameworks that allow accurate reconstruction of scene properties using small numbers of photons. These frameworks first model the statistics of individual photon detections, which are observations of an inhomogeneous Poisson process, and express a priori scene constraints for the specific imaging problem. Each yields an inverse problem that can be accurately solved using novel variations on sparse signal pursuit methods and regularized convex optimization techniques. We demonstrate our frameworks' photon efficiencies in six imaging scenarios that have been well-studied in the classical settings with large numbers of photon detections: single-depth imaging, multi-depth imaging, array-based timeresolved imaging, super-resolution imaging, single-pixel imaging, and fluorescence imaging. Using simulations and experimental datasets, we show that our frameworks outperform conventional imagers that use more naive observation models based on high light-level assumptions. For example, when imaging depth, reflectivity, or fluorescence lifetime, our implementation gives accurate reconstruction results even when the average number of detected signal photons at a pixel is less than 1, in the presence of extraneous background light. |
Using connectivity to build e-Government | Developed countries are increasingly using information technology to access and deliver information across all the organizations that comprise the government. Argentina is in the first stages of adopting Internet technology for the government. This thesis proposes that Argentina's e-government construction can be optimized if a good foundation for systems connectivity is put in place first. The author argues that the current time lag in the construction of Argentina's e-Government can be used in a positive way if experience from more developed countries is used. Chapter 1 provides a strategy for building e-Government in Argentina using standardization and connectivity. In Chapter 2, a theoretical analysis about the economics of systems integration is presented. In Chapter 3 the importance of standardization and specially the use of XML are outlined. Chapter 4 develops the basis of IT infrastructure and its importance to enable e- Government. Chapter 5 evaluates the current state of e-Government in three developed countries. Chapter 6 and 7 provide a potential framework for building the e-Government in Argentina based in the ideas of the previous chapters. The author has found that developed countries are defining standard connectivity based on XML and that they are in the process of becoming a full e-government by 2005. At the same time Argentina is lacking a coherent plan to enable the creation of the e-Government in the near future. Finally, the thesis proposes the standardization of systems interfaces using XML and the definition of standard XML data schemes to facilitate connectivity. At the same time the thesis encourages building an IT infrastructure and the use of standard applications to gain efficiency and lower the cost of e-Government. |
Study of base isolation systems | The primary objective of this investigation is to outline the relevant issues concerning the conceptual design of base isolated structures. A 90 feet high, 6 stories tall, moment steel frame structure with tension cross bracing is used to compare the response of both fixed base and base isolated schemes to severe earthquake excitations. Techniques for modeling the superstructure and the isolation system are also described. Elastic time-history analyses were carried out using comprehensive finite element structural analysis software package SAP200. Time history analysis was conducted for the 1940 El Centro earthquake. Response spectrum analysis was employed to investigate the effects of earthquake loading on the structure. In addition, the building lateral system was designed using the matrix stiffness calibration method and modal analysis was employed to compare the intended period of the structure with the results from computer simulations. Base isolation proves to be effective in reducing the induced inertia forces on a structure by increasing the effective period of oscillation. Keywords: Base Isolation, time history analysis, response spectrum analysis, matrix stiffness calibration method. |
Measuring high-performance embedded systems | The Extreme Benchmark Suite (XBS) is designed to support performance measurement of highly parallel "extreme" processors, many of which are designed to replace custom hardware implementations. XBS is designed to avoid many of the problems that occur when using existing benchmark suites with nonstandard and experimental architectures. In particular, XBS is intended to provide a fair comparison of a wide range of architectures, from general-purpose processors to hard-wired ASIC implementations. XBS has a clean modular structure to reduce porting effort, and is designed to be usable with slow cycle-accurate simulators. This work presents the motivation for the creation of XBS and describes in detail the XBS framework. Several benchmarks implemented with this framework are discussed, and these benchmarks are used to compare a standard platform, an experimental architecture, and custom hardware. |
Characterization of synchronizer performance for a clutchless transmission | Synchronizers are a ubiquitous component of almost every type of transmission in modem vehicles. They are mechanical devices whose function is to ensure that components rotating at different rates can be harmonized smoothly and without eroding their surfaces. They are responsible for both the durability of the transmission and the comfort of the passengers. This work analyzes the capabilities and limitations of synchronizers to be used in a novel transmission. It is a contribution to a larger project whose goal is to develop a hybrid, clutchless transmission for a performance vehicle that will improve efficiency by eliminating the friction and mechanical losses inherent in a traditional clutch. An overview of the synchronization process is presented followed by a simplified mathematical model of the common baulk-ring synchronizer. The model is experimentally validated in order to make predictions of the device's performance on the new transmission. Several simulated scenarios are then developed that provide information that is critical for designing synchronizers for the clutchless transmission. Matlab code was developed for these simulations and is provided at the end for replication of the results. Considering the demanding environment under which the synchronizers are expected to operate in the clutchless transmission, the possible failure modes of the synchronizer components are investigated. Finite element analysis (FEA) is used to predict the maximum loads on the synchronizer ring before the material yields. An energy analysis is also performed to ensure that the energy dissipation rate of the friction surfaces is adequate. |
Segmentation and visualization for cardiac ablation procedures | In this thesis, we present novel medical image analysis methods to improve planning and outcome evaluation of cardiac ablation procedures. Cardiac ablation is a common medical procedure that consists of burning cardiac tissue causing atrial fibrillation, or irregular contractions of the heart's atria. We first propose a method for the automatic delineation of the left atrium in magnetic resonance (MR) images acquired during the procedure. The high anatomical variability of the left atrium shape and of the pulmonary veins that drain into it presents significant difficulties for cardiac ablation. Consequently, accurate visualization of the patient's left atrium promises to substantially improve intervention planning. We perform the segmentation using an automatic atlas-based method, which makes use of a set of example MR images in which the left atrium was manually delineated by an expert. We demonstrate that our approach provides accurate segmentations that are also robust to the high anatomical variability of the left atrium, while outperforming other comparable methods. We then present an approach to use this knowledge of the shape of the left atrium to aid in the subsequent automatic visualization of the ablation scars that result from the procedure and are visible in MR images acquired after the surgery. We first transfer the left atrium segmentation by aligning the pre and post-procedure scans of the same patient. We then project image intensities onto this automatically generated left atrium surface. We demonstrate that these projections correlate well with expert manual delineations. The goal of the visualization is to allow for inspection of the scar and improve prediction of the outcome of a procedure. This work has a potential to reduce the considerable recurrence rates that plague today's cardiac ablation procedures. |
Mapping bulk electrical properties with non-contact RF measurements | The human body is composed primarily of dielectric tissue with spatially varying permittivity and conductivity. Traditional MRI does not measure these properties. Instead, the conductivity of the patient is a nuisance, causing unpredictable detuning of coils and field inhomogeneities. This thesis presents a method for mapping the electrodynamic properties of the patient's body with both MR and non-MR techniques. Such mapping has direct applications for medical imaging and SAR calculation. |
Generation gaps in engineering? | There is much enthusiastic debate on the topic of generation gaps in the workplace today; what the generational differences are, how to address the apparent challenges, and if the generations themselves are even real. Despite the frenzy, however, there has been relatively limited empirical research performed on this issue; in particular, no empirical research in the context of an engineering organization seems to have been performed. With that in mind, the main intent of the thesis is to determine the existence of generational differences among engineers in a large, technical organization, and discuss the potential implications of the findings. Extensive literature research and analysis identified communication techniques, leadership preferences, learning styles, and work motivation to be the key hypothesized differences most pertinent to an engineering organization. A questionnaire was put together based on these differences and distributed to a sample that consisted of approximately eighty members from each of the four generations: Traditionalists, Baby Boomers, Generation Xers, Millennials. In general, data analysis revealed that engineers do not exhibit generation gaps to the same extent as described in the literature. In fact, much of the data suggested contrary views, most notable of which was the importance of face-to-face communication, especially amongst the younger engineers. Knowledge transfer across generations remains a challenge, however, and possible approaches to addressing this issue are through physical accommodations, extensive mentorship programs, and salient investments in Information Technology (IT). |
Analysis of traditional Korean space and its application to a contemporary problem : a crematorium | This is more of an experimental design thesis rather than a theoretical research-oriented thesis on the history of Korean Architecture. Ultimately, a design of a concrete building, a crematorium, will be the final product of this thesis and will serve as a vehicle with which to experiment with my ideas in the design process. There is a strong driving force in this thesis: the aspiration to interpret the architecture of my ancestors with a fresh eye - in light of spatial organization. Thus, it was a necessary first step to visit a few historical buildings to experience the spaces themselves, following with an analysis of the plan afterwards. Throughout my own study, some assumptions were drawn as design guid~lines for the design of a crematorium. The critical issue of this design thesis centers on the evolution of 'Non-diagrammatic Space' from 'Diagrammatic Space' in the history of Korean architecture. Designing a building has been thought a complex process which relates political, economical, and social factors at the same time. But, though we know this intuitively, a design methodologist recently proved that most designs are generated not by listing all constraints but by starting with a 'prime generator'. In this context. the diagrammatic Korean temples are good illustrations of the 'Prime Generator'. But my main concern lies in ideas that cannot be easily explained by the notion of prime generator in the non-diagrammatic temples, which were the outgrowth of deep contemplation and a process of trial and error through hundreds of years. |
Contribution to the culture of "college thinking" | Four fragments constitute the focus of analysis of this study; at the same time they are used as objects in the experimental set up of the thesis. This set up investigates the possibilities of a collage as a model for a research. The Vitruvian notion of the architect as a person standing between a variety of subject fields is connected to Kurt Schwitters collage techniques. The four pieces researched with four different research modes: Art, scientific research, report and design: 1. A collection of images and collages. This collection consists of images of snails, Franciscan architecture, personal memory and allegorical objects. 2. A historic research of the initial state of the Franciscan Movement and St. Francis with respect to their attitude towards buildings. 3. Three case studies of Franciscan Facilities in Boston and Providence. 4. A design proposal for an extension of the homeless day care center St. Francis House, which is located at the Boston Common. |
The impact of pathological ventilation on aerosol deposition : imaging, insight and intervention | Aerosol therapies are often used to treat lung diseases in which ventilation is distributed heterogeneously throughout the lung. As therapeutic aerosols are transported by the inhaled air, it is likely that deposition is diminished within poorly ventilated regions of the lung. These regions are often the most in need of therapy. We measured the effects of heterogeneous ventilation on aerosol deposition in a group of bronchoconstricted asthmatic subjects. We then developed a new image processing technique which allowed us to identify the anatomical location of aerosol deposition. This technique accounted for blurring due to limited resolution of the PET image, motion artifacts due to breathing, and registration uncertainty. We introduced a theoretical framework to characterize four mechanisms of variability in deposition between peripheral regions of the lung. This framework added insight into the interaction between ventilation and deposition, and will permit the future comparison of the experimental data with computational models. Together, the imaging data and theoretical framework suggested that more than a third of the observed variability in the deposition per unit volume among lung lobes was due to heterogeneous ventilation. Using helium-oxygen as a carrier gas for aerosol has been considered as a potential intervention to homogenize deposition in the lung periphery. To investigate this, we repeated the PET-CT measurements in a second group of bronchoconstricted asthmatic subjects breathing helium-oxygen, and compared the results to those of the group breathing room air. We did not find systematic differences in the deposition patterns of the two groups, although the relationship between ventilation and aerosol deposition tended to be stronger in the group that used helium-oxygen as the carrier gas. Finally, we used analytical tools and an in-silico model of bronchoconstriction to illustrate the emergence of pendelluft gas transport between parallel regions of the lung. We found that though pendelluft may emerge in asthma, the overall volume passed between parallel regions of the lung is likely less than 2% of the tidal volume, and thus is not likely to substantially influence aerosol deposition. |
Utilizing economic theories of retail to revitalize inner-city neighborhood business districts : the case of Uphams Corner Main Street | Community development corporations, as well as other stakeholder organizations in inner-city areas, are increasingly taking on the tasks of economic development in general and commercial revitalization in particular. In order to improve the effectiveness of these interventions, this thesis has two purposes: 1) to provide a rigorous, widely applicable framework for approaching the task of revitalizing inner-city neighborhood business districts, and 2) to provide an example of how this framework can be used by applying it to the Uphams Corner neighborhood in Dorchester, Massachusetts. Four economic theories of retail provide the backbone of this framework. Economists use these theories to explain the success of malls and to describe what attracts people to one shopping area over another. This thesis also uses the theories to explore the strategy of orienting commercial districts around particular niches or clusters of stores. While many have recommended niche-based strategies, they do not explicitly connect this to the economic dynamics of retail areas. Therefore, this thesis seeks to explain: 1) what creates the connection between stores in various types of niches, 2) how the nature of competition between stores differs across various types of niches, 3) how these theories can be used to develop concrete strategies for working with existing stores and recruiting new ones, and 4) which types of new stores will be the most successful and add the most to the health of the area. In the process of developing this framework, the thesis describes a number of doubts as to whether the economic theories of retail are applicable to inner city areas. If these doubts can not be overcome, it does not make sense to use the theories in these districts. The first doubt concerns whether or not commercial revitalization in general, and the economic theories in particular, can be used to achieve the goals of community economic development. The second is that improving factors such as parking, cleanliness, and safety may be more important to neighborhood commercial revitalization than altering the economics of the area. The third is that malls may only be able to take advantage of these economic theories because they are owned by a single entity. Inner-city commercial districts have multiple property owners and under resourced store owners, making coordination more difficult. The fourth is that there may be something different about inner-city consumers such that they might not shop in the way that the theories predict. This thesis argues that each of these doubts can be overcome, and that the theories can be used effectively to revitalize inner-city neighborhood business districts. The thesis concludes with the argument that more emphasis should be placed on economically based strategies for revitalizing commercial areas. The current models, the Main Street approach, Business Improvement Districts, and others focus primarily on improving non-economic, physical factors of these areas. While these efforts are important, this thesis asserts that a greater understanding of the economic dynamics of these retail nodes should permeate all of the work of these organizations. |
Analyzing the impact of transportation changes on Boston's neighborhoods | The Federal Aid Highway Act of 1956 provided the funds for Massachusetts to build a system of highways that threatened to cut through various neighborhoods in the greater Boston area. A broad coalition of people put a stop to these plans in 1972, and advocacy at the federal level allowed for highway funds to be shifted to public transportation projects. Many organizations within this coalition had broader goals of improving the livelihoods of their communities beyond just shutting down highways. Since equitable public transportation can play a key role in improving economic mobility, this thesis explores how changes in public transit stemming from the Boston anti-highway movement impacted nearby neighborhoods and assesses the areas that may still be lacking in access to adequate transit today. |
Genomic studies of motif enrichment and conservation in the regulation of gene expression in the brain | Several bioinformatic tools will be brought to bear in this thesis to identify specific genomic loci that serve as regulatory gateways of gene expression in brain. These "motifs" are short nucleotide patterns that occur in promoters and 5' or 3' untranslated regions of genes. Occurrences of motifs that function in eukaryotic genomes as, e.g., transcription factor binding sites or targets of RNA interference are assumed to lie at the nexus of several trends. Instances that are indeed regulatory and not just bits of random sequence should show evidence of actual binding of factors that have a significant effect on expression levels. Such motif instances are also expected to be significantly enriched (or de-enriched), compared to background, in the genes regulated by their binding factors and in brain structures most closely associated with these genes' functions. Finally, truly regulatory motif instances are likely to be highly conserved in orthologous genes across multiple genomes; i.e., conservation can be taken as a proxy for function. My research exploits these ideas by exploring genome-wide properties of motifs associated with the transcription factor family MEF2, some of whose members are known to play a role in synapse development. Data from chromatin immunoprecipitation and tiling-microarray (ChIP-on-chip) experiments [11 have isolated peaks of specific binding by MEF2 in developing rat brains. Conservation and enrichment of these sites are analyzed here for their association with functionality and variability of motifs in genes that have been shown to fall under the control of MEF2 in excitatory neurons. |
Design and control of a planar two-link manipulator for educational use | This paper proposes a new robotic planar two-link manipulator design for educational use. Planar two-link manipulators are among the most accessible two-degree-of-freedom robots for students because they function like human arms. As a result they are ideal for laboratory teaching environments. While previous designs using belt-driven arms served adequately, this new design possesses a number of features that were not possible with the previous design, including more intuitive simplified dynamics, an expanded workspace allowing multiple full rotations, and the ability to be easily reconfigured into an acrobot (an underactuated double-pendulum which can be stabilized in a vertical configuration while being actuated only at the middle joint). The governing equations of the system are derived and an analysis of velocity control in the xy plane is perform and a control methodology is also presented by which the arm can be stabilized vertically while in its acrobot configuration. A Discussion of tradeoffs relevant to the future design of similar systems is also presented. |
Force-modulated growth of CNT forests | Carbon nanotubes (CNTs) can have exceptional mechanical, thermal, and electrical behavior, but successful use of CNTs in macroscale materials requires scalable processes to organize CNTs while preserving their intrinsic properties. Although the chemical influences on CNT growth have been subject to significant research, the effect of mechanical forces, which have been shown to influence the kinetics of many chemical reactions, remains unclear. When CNTs grow into vertically aligned "forest" structures by chemical vapor deposition (CVD), entanglement and attractive forces among the CNTs cause the CNTs to become mechanically coupled. Further, because there is a distribution of sizes, orientations, and growth rates amongst individual CNTs within a forest, it has been suggested that coupled CNTs develop forces that are transmitted to the growth interface at the catalyst. The goal of this thesis is to investigate the effects of applied mechanical forces and intrinsic forces that result from mechanical coupling on CNT growth. Using a custom-built chemical vapor deposition system including a micromanipulator that can apply compressive loads to a CNT forest and measure its height in real-time, the effects of extrinsic mechanical forces on the growth kinetics and forest morphology were studied. It was found that forces as small as 0.1 nN per CNT decrease the collective growth rate of the CNT forest along with its terminal height. Furthermore, time-varying forces can be used to induce a morphological change in the forest structure and modulate the apparent growth rate without causing termination. Next, a finite element model was developed to simulate the forces exerted between a pair of CNTs growing at different rates and coupled by van der Waals forces. The simulation predicts that mechanical coupling between CNTs enables forces exceeding 10 nN to be transmitted to the catalyst, which are potentially several orders of magnitude larger than the externally applied forces that were found experimentally to influence forest growth. Together, these findings suggest that the quality and growth rate of CNT forests may be limited by CNT-CNT mechanical coupling and force transmission to the catalyst, and motivate future work using controlled forces to manipulate the quality and morphology of CNTs for various applications. |
Thermal modulation during solvent annealing of PS-PDMS block copolymer | The self-assembly of block copolymers (BCP) has been a promising area of research for nanolithography applications in microelectronics because of their ability to produce nano-scale level periodic structures with long-range order. Ideal BCPs for generating these nano-scale patterns fall within the strong segregation limit (SSL) and have a high interaction parameter to drive BCP phase transitions. BCP morphologies can vary from equilibrium structures such as spheres, cylinders, and gyroid, to metastable structures such as hexagonal perforated lamellar (HPL). A variety of processing techniques including solvent vapor annealing (SVA) have been developed in order to facilitate the phase transitions of BCPs from disordered to ordered states. SVA parameters which can affect the final film morphology include the swelling thickness of the film and solvent removal rate. Thermal modulation of the substrate was used to explore the effects of rapid solvent evaporation during the annealing process on the morphologies of the PS₁₆-b-PDMS₃₇ system. Additional cycles of solvent update and film reswelling were introduced into the annealing procedure to induce greater long-range ordering of film morphologies. Although a range of morphologies were explored, there was special focus on developing a procedure for mono-layer HPL structures for nanolithography applications. |
Off-line calibration of DTA models | Advances in Intelligent Transportation Systems (ITS) have resulted in the deployment of surveillance systems that automatically collect and store extensive network-wide traffic data. Dynamic Traffic Assignment (DTA) models have also been developed for a variety of dynamic traffic management applications. Such models are designed to estimate and predict the evolution of congestion through detailed models and algorithms that capture travel demand, network supply and their complex interactions. The availability of rich time-varying traffic data spanning multiple days thus provides the opportunity to calibrate a DTA model's many inputs and parameters, so that its outputs reflect field conditions. The current state of the art of DTA model calibration is a sequential approach, in which supply model calibration (assuming known demand inputs) is followed by demand calibration with fixed supply parameters. In this thesis, we develop an off-line DTA model calibration methodology for the simultaneous estimation of all demand and supply inputs and parameters, using sensor data. We adopt a minimization formulation that can use any general traffic data, and present approaches to solve the complex, non-linear, stochastic optimization problem. |
Of energy and the economy : theory and evidence for their functional relationship | This paper offers a set of explicit functional relationships that link energy and the economy. Despite the reliance on energy permeating the whole economy, no such complete relationships had been presented before. How related are energy and the economy? What role does energy play in the economic growth? Motivated to seek an explicit functional answer, I theorize the role of energy and then test it with economic models, using data for 16 OECD countries from 1980 to 2001. First, I find that energy is a cross-country representative good whose prices are equalized when converted to a reference currency. Thus, energy prices satisfy the purchasing power parity. For all but one country, the half life of the real energy exchange rate is less than a year and as low as six months, shorter than those derived by other real exchange rate measures. Second, considering energy a cross-time representative good, I obtain that a country's utility function is inversely proportional to both its income share of energy and its energy price. I also obtain an explicit, unified two-dimensional (cross countries and time) production function with energy and non-energy as the two inputs. Third, I conclude a cross-country parity relationship for income shares of energy, similar to that for energy prices. |
The role of pyruvate kinase regulation in tumor growth and metabolism | Cancer is a disease of inappropriate cell proliferation, and central carbon metabolism is highly regulated to support the unique anabolic needs of proliferating cells. Pyruvate kinase, the enzyme catalyzing the final step of glycolysis, is an important point of regulation. Mammals have four pyruvate kinase isoforms, and one in particular, the M2 isoform, is preferentially expressed in proliferative tissues, including cancers. We sought to determine the role and importance of PKM2 in cancer. The activity of pyruvate kinase M2 (PKM2) is down-regulated by pro-proliferative signaling in the cell, and reduced PK activity appears to be important for proliferative metabolism. Because oncogenic signaling reduces PKM2 activity, we hypothesized that artificially high intracellular pyruvate kinase activity would disrupt proliferative metabolism and hinder tumor growth. We found that increased PK activity due to expression of the constitutively-active PKM1 isoform or direct pharmacological activation of PKM2 perturbs metabolism and reduces tumor growth in a lung cancer xenograft model. These results suggested that PKM2 expression is selected for in cancers because PKM2 activity can be down-regulated in a controlled fashion. We next sought to determine if the PKM2 isoform is necessary for tumor proliferation. Deletion of PKM2 in a mouse model of BRCA1-deficient breast cancer demonstrated that PKM2 is not required for tumor formation or growth. PKM2-null tumors exhibited heterogeneous PKM1 expression, and tumor cell proliferation was associated with low PKM1 expression in the absence of PKM2. Analysis of human breast tumors revealed highly variable PKM2 protein expression, and heterozygous PKM2 mutations were found in many cancer types. These mutations cause truncations or amino acid changes in conserved regions of the enzyme. Determination of kinetic parameters of purified wild-type and mutant PKM2 showed that cancer-associated missense mutations reduce affinity of the enzyme for substrate, reduce maximum velocity, or disrupt response of the enzyme to activation by its allosteric activator FBP. These results suggest that cancer cells tolerate or select for reduced pyruvate kinase activity. We conclude that PKM2 is important because its activity is down-regulated to support proliferation, but the PKM2 isoform itself is not required for tumor formation or cancer cell proliferation. |
High fidelity quantum gates with ions in cryogenic microfabricated ion traps | While quantum information processing offers a tantalizing possibility of a significant speedup in execution of certain algorithms, as well as enabling previously unmanageable simulations of large quantum systems, it remains extremely difficult to realize experimentally. Recently, fundamental building blocks of a quantum computer, including one and two qubit gates, teleportation and error correction, were demonstrated using trapped atomic ions. Scaling to a larger number of qubits requires miniaturization of the ion traps, currently limited by the sharply increasing motional state decoherence at sub-100 [mu]m ion-electrode distances. This thesis explores the source and suppression of this decoherence at cryogenic temperatures, and demonstrates fundamental logic gates in a surface electrode ion trap. Construction of the apparatus requires the development of a number of experimental techniques. Design, numerical simulation and implementation of a surface electrode ion trap is presented. Cryogenic cooling of the trap to near 4 K is accomplished by contact with a bath cryostat. Ions are loaded by ablation or photoionization, both of which are characterized in terms of generated stray fields and heat load. The bulk of new experimental results deals with measurements of electric field noise at the ion's position. Upon cooling to 6 K, the measured rates are suppressed by up to 7 orders of magnitude, more than two orders of magnitude below previously published data for similarly sized traps operated at room temperature. The observed noise depends strongly on fabrication process, which suggests further improvements are possible. The measured dependence of the electric field noise on temperature is inconsistent with published models, and can be explained using a continuous spectrum of activated fluctuators. The fabricated surface electrode traps are used to demonstrate coherent operations and the classical control required for trapped ion quantum computation. The necessary spectral properties of coherent light sources are achieved with a novel design using optical feedback to a triangular, medium finesse, cavity, followed by electronic feedback to an ultra-high finesse reference cavity. |
An investigation into and recommendations for appropriate technology education | The purpose of this paper is to present an example of appropriate technology (AT) education in a university setting, and make recommendations for using open source technology to aid AT education (OSAT). This paper presents a brief overview of the AT movement, and defines critical criteria for creating and implementation solutions for the developing world using this approach. The International Development Initiative (IDI) at the Massachusetts Institute of Technology is described in detail as a model example of efforts to promote the study of AT in higher education. OSAT is investigated in further detail to prove validity as a new aspect of the AT movement and a course of study incorporating the use of OSAT is developed to aid educators. |
Functional and computational analysis of RNA-binding proteins and their roles in cancer | This work is concerned with mRNA processing in mammalian cells and proceeds in two parts. In the first part, I introduce a computational framework for inferring the abundances of mRNA isoforms using high-throughput RNA sequencing data. This framework was applied to study the targets of the ubiquitous splicing factor hnRNP H in human cells. In the second part, I describe an experimental study of the Musashi (hnRNP-like) family of RNA-binding proteins in stem cells and cancer cells, which incorporates computational analyses that rely heavily on the framework developed in part one. In sum, this work provides a computational framework of general use in global analyses of RNA processing and its protein regulators, as well as functional insights into a family of poorly understood RNA-binding proteins. Several related analyses and techniques developed as part of the thesis are described in Appendix A-C. Appendix A describes a study of activity-dependent gene expression and mRNA processing in the mouse olfactory bulb. It uses computational techniques developed in part one of the thesis. Appendix B describes a technique for quantitative visualization of alternative splicing from RNA sequencing data and its integration into a genome browser. Appendix C describes a method for clonal analysis of neural stem cell growth and differentiation in culture using live imaging and `microdot' plates, developed as part of the work presented in part one of the thesis. |
New storage architecture for distributed storage systems | This thesis introduces a new NAND flash-based storage architecture, NOHOST, for distributed storage systems. A conventional flash-based storage system is composed of a number of high-performance x86 Xeon servers, and each server hosts 10 to 30 solid state drives (SSDs) that use NAND flash memory. This setup not only consumes considerable power due to the nature of Xeon processors, but it also occupies a huge physical space compared to small flash drives. By eliminating costly host servers, the suggested architecture uses NOHOST nodes instead, each of which is a low-power embedded system that forms a cluster of distributed key-value store. This is done by refactoring deep I/O layers in the current design so that refactored layers are light-weight enough to run seamlessly on resource constrained environments. The NOHOST node is a full-fledged storage node, composed of a distributed service frontend, key-value store engine, device driver, hardware flash translation layer, flash controller and NAND flash chips. To prove the concept of this idea, a prototype of two NOHOST nodes has been implemented on Xilinx Zynq ZC706 boards and custom flash boards in this work. NOHOST is expected to use half the power and one-third the physical space as compared to a Xeon-based system. NOHOST is expected to support the through of 2.8 GB/s which is comparable to contemporary storage architectures. |
Mixed-use design in a Cape Cod harbor | Change is an ongoing, unavoidable process that may occur for many reasons and in many ways. In this world of uncertainty and often unwanted surprises it is comforting to know that there are some aspects of our physical and social environment that remain constant while others are changing. This thesis examines various attitudes towards new and old, towards continuity and change. It then centers on the idea of transformation which acknowledges the necessity of change while still respecting the positive aspects of what exists. This idea of transformation is demonstrated in general through diagrams, and specifically through a design of a mixed-use development in the harbor at Woods Hole on Cape Cod. |
Tactile sensing of shape : biomechanics of contact investigated using imaging and modeling | The overall goal of this research effort is to improve the understanding of the biomechanics of skin as it pertains to human tactile sense. During touch, mechanoreceptors beneath the skin surface are mechanically loaded due to physical contact of the skin with an object and respond with a series of neural impulses. This neural population response is decoded by the central nervous system to result in tactile perception of properties such as the shape, surface texture and softness of the object. The particular approach taken in this research is to develop a realistic model of the human fingertip based on empirical measurements of in vivo geometric and material properties of skin layers, so that the mechanical response of the fingertip skin to different shapes of objects in contact can be investigated, to help identify the relevant mechanism that triggers the mechanoreceptors in tactile encoding of object shape. To obtain geometric data on the ridged skin surface and the layers underneath together with their deformation patterns, optical coherence tomography (OCT) was used to image human fingertips in vivo, free of load as well as when loaded with rigid indenters of different shapes. |
Regional architectures and environmentally-based transportation planning : an institutional analysis of planning in the Mexico City Metropolitan Area | This thesis presents research motivated by three critical issues. First, the rapid penetration of information technologies has changed the face of both transportation system management and long-term planning. Second, the federal mandate for integrated environmentally-based transportation planning in the United States was expanded with firm resolve by the Intermodal Surface Transportation Efficiency Act (ISTEA). Third, ongoing research work on the air quality problem in Mexico City demands in-depth attention to the management and planning of that metropolitan region's transportation system because of the important contribution of mobile sources to the air pollution problem. The purpose of this thesis is to expand an established theoretical framework that addresses the issue of data-intensive integrated management to include the relationship between transportation and environmental planning. The theoretical product of that endeavor, namely the Regional Planning Architecture framework, is tested in the context of Mexico City, in which the mobility and environmental problems are more extensive than in any city in the United States. The result of this work is a new Regional Planning Architecture (RPA) framework to accompany the previously established Regional Service Architecture. The RPA describes the fundamental institutional relationships that surround the production and implementation of short- and longterm transportation plans. The architecture reflects the need to integrate transportation and environmental plans by incorporating the relationships between transportation and environmental planners. This is particularly important in the area of conformity planning, which was brought about by ISTEA in 1991. Initially based on the specific requirements of U.S. legislation, the use of the Mexico City case study leads to an independent platform that helps to produce original recommendations for improvement in that city's mobility and environmental systems. The analysis suggests that in using regional architectures as a diagnostic or prescriptive tool, one should emphasize five elements of interaction among institutions: goals, ideas/needs, funding, approval, and data. The focus on goals reflects the finding that the formulation and exchange of goal statements can help planners comprehend and incorporate the goals of their colleagues from other agencies. Funding and emissions constraints facilitate the application of goals in the prioritization/approval process. The sharing of data and ideas/needs reflects the increasing availability of information to planners on the effectiveness of prior strategies and investments. |
Vulnerability analysis of electricity distribution networks with large-penetrations of PEVs and DERs | This thesis focuses on the vulnerability assessment of radial electricity distribution networks (DNs) under large-scale integration of Distributed Energy Resources (DERs) and Plug-in Electric Vehicles (PEVs). We formulate a two-player Stackelberg security game involving an attacker (external threat agent) and the defender (network operator). First, the attacker targets a subset of the insecure DER or PEV nodes, and strategically manipulates their set-points by attacking the DER/PEV controller logic at the nodes. Next, the defender responds to the resulting supply-demand mismatch by triggering network control operations, which includes direct load control and control of available non-compromised DERs/PEVs. The attacker's (resp. defender's) objective is to maximize (resp. minimize) the weighted sum of the cost of active and reactive power supply, costs of DER/PEV and load control, and the cost due to loss of voltage regulation. This composite cost captures the key trade-offs that the network operator faces in balancing power supply and quality during a broad range of contingency conditions. The choice of this cost in the security game reflects the attacker's overall goal of comprising the DER/PEV nodes to maximize the minimum composite cost for the network operator. Solving the sequential game with nonlinear power flow constraints is a computationally hard problem. To address this challenge, we introduce two auxiliary sequential game problems each with linear power flow constraints. We prove that the values of these relaxed problems upper and lower bound the value of the original game. Next, we introduce a greedy algorithm that can be utilized to efficiently compute an optimal attack strategy for both auxiliary games. Our main result is that, under a set of assumptions, the set of optimal attacker strategies is identical for these games, and hence we obtain a tractable solution to compute an optimal attack for the original game. Furthermore, the optimal attack strategy exhibits an interesting structural property: the downstream nodes are more critical for limiting costs of reactive power supply and maintaining voltage regulation. This insight is useful for vulnerability assessment of DNs under DER/PEV node compromises. Finally, we also exploit the structure of optimal attack to design a distributed control strategy for defender response. |
Plant Homeodomain Finger Protein 6 modulates the chromatin landscape in B-cell leukemia | Developmental and lineage plasticity have been observed in numerous malignancies, and have been correlated with tumor progression and drug resistance. However, little is known about the molecular mechanisms that enable such plasticity to occur. Here, we describe the function of the Plant Homeodomain Finger Protein 6 (PHF6) in leukemia and define its role in regulating chromatin accessibility to lineage-specific transcription factors. We show that loss of Phf6 in B-cell leukemia results in systematic changes in gene expression via alteration of the chromatin landscape at the transcriptional start sites of B- and T-cell specific factors. Additionally, Phf6KO cells show significant down-regulation of genes involved in the development and function of normal B-cells, up-regulation of genes involved in T-cell signaling, and give rise to mixed-lineage lymphoma in vivo. Engagement of divergent transcriptional programs results in phenotypic plasticity that leads to altered disease presentation in vivo, tolerance of aberrant oncogenic signaling, and differential sensitivity to frontline and targeted therapies. These findings suggest that active maintenance of a precise chromatin landscape is essential for sustaining proper leukemia cell identity, and that loss of a single factor (PHF6) can cause focal changes in chromatin accessibility and nucleosome positioning that render cells susceptible to lineage transition. |
Novelty in goal-oriented machines using a thread memory structure | Resourcefulness and creativity are desirable properties for an intelligent machine. The incredible adeptness of the human mind at seeing situations from diverse viewpoints allows it to conjure many techniques to accomplish the same goal, and hence recover elegantly when one method fails. In the context of goal-oriented machines, this thesis presents a system that finds substitutes for the typical physical resource used to accomplish a goal, by finding novel uses for other, available resources-uses that these resources were not originally meant or designed for. In a domain where an object can serve multiple functions, this requires: (1) understanding the functional context the object is operating in; (2) building a realistic representation of the given objects, which often do not fall neatly into tightly-structured categorizations, but instead share properties with other 'boundary' objects. The system does this by learning from examples, and using the average member, or 'stereotype' as the class representative; (3) allowing imperfection: identifying properties that are not crucial for goal satisfaction, and selectively ignoring them; and (4) measuring similarity between objects to find the best substitute. The system bootstraps with knowledge about the properties of the objects and is given positive and negative examples for the goal. It can infer, for example, that two objects such as an orange (the typical resource) and a ball (the positive example) are related in the context of finding a throwable object on account of their similarity in shape and size, but unrelated in the context of finding an ingredient for a fruit salad, because one is a fruit and the other is not. |
The engineers and the urban system, 1968-1974/ | The social policy agenda of the Great Society was shaped by the recognition that if broad social improvement was to be achieved in urban America, social planning and state intervention based on systemically acquired expertise about the city would have to be developed. As a case study in the development of such expertise, in this thesis I explore the work of the Urban Systems Laboratory (USL), established in 1968 at the Massachusetts Institute of Technology (MIT) with funding from the Ford Foundation. Using computationally intensive methods, behavioral models and the latest techniques of the information sciences, research at the USL emphasized the role of rational, analytical, social scientific expertise in managing social conflict. In particular, I explore the work of Jay W. Forrester, a member of the USL whose research on the city was published in 1969 as Urban Dynamics. Using an IBM Systems/360 Model 67, Forrester built one of the first large-scale, interactive, computational models of a city specifically to explore the consequences of the social policies of the period and more generally the possibility of the social engineering of complex social systems in a postindustrial society. This project of the production of expertise at the USL struggled to secure legitimacy in the early 1970s as the attempt to treat the problems of urban America as phenomena to be handled by a new class of experts was overwhelmed by the sheer scale of urban turmoil. |
Fast scheduling for OFS | Optical Flow Switching (OFS) is a key enabler of future scalable optical networks. In the past decade, the OFS architecture has been studied to build an all-optical data plane to provide an end-to-end, cost-effective data transport to users with large transactions. Flow switching provides low-cost service to high-end users by relieving the IP routers on the edge of wide area networks from large transactions. However, the scheduling process of OFS presents possible queuing delays of several transaction durations. For some special applications with urgent time deadlines, the users want to bypass the queuing and pay more to use the network as soon as possible. In this thesis, we propose a fast scheduling algorithm which utilizes a probing approach to enable OFS to set up end-to-end connections for users with urgent transactions with a delay of slightly more than one round-trip time. A central control manager is used to periodically collect from network regions their most recent entropy evolutions of the network states and broadcast this information across the whole network in the control plane. With this information, fast setups of end-to-end all-optical connections for OFS are achieved by probing independent paths between source and destination, and reserving the available light paths along the way. A modified Bellman-Ford algorithm is designed to select the paths with the least blocking probabilities. By grouping details of network states into the average entropy, we can greatly reduce the information collected and disseminated by the centralized controller, making the network management and control system scalable to large networks. Since our algorithm makes no assumptions about network models or traffic statistics, it is robust against model variations, and any future changes in network topologies and traffic patterns. The technique can also be used in heterogeneous networks, in which networks from different domains are interconnected to provide a broader coverage. |
Design and fabrication of a maneuverable robot for in-pipe leak detection | Leaks in pipelines have been causing a significant amount of financial losses and serious damages to the community and the environment. The recent development of in-pipe leak detection technologies at Massachusetts Institute of Technology made it possible to find the accurate location of leaks in underground pipes. However, like all in-pipe leak sensors, they need a maneuverable robot to transport them inside the pipes. In the pipe networks, the robot must be able to perform complicated movements such as sharp turns at Tee junctions. This thesis presents a solution to this in-pipe leak detection challenge and the design and a prototype of such robot. The design and fabrication of a small in-pipe swimming robot of high maneuverability is presented. The robot is equipped with a pair of customized micro RIM driven propellers which provide a powerful and safe propulsion. A prototype robot that operates in 10 cm diameter pipes is built and tested experimentally. This robot demonstrated experimentally abilities to follow straight lines and make turns with radii smaller than a fraction of its body length. |
Fluorogenic probes for live-cell imaging of biomolecules | Fluorogenic probes, small-molecule sensors that unmask brilliant fluorescence upon exposure to specific stimuli, are essential tools for chemical biology. Probes that detect enzymatic activity can be used to illuminate the complex dynamics of biological processes at a level of spatiotemporal detail and sensitivity unmatched by other techniques. This dissertation describes the development of new fluorophore chemistries to expand our current fluorogenic probe toolkit and the subsequent application of these probes to study dynamic cell transport processes. Chapter 1. Enzyme-Activated Fluorogenic Probes for Live-Cell and In Vivo Imaging. Chapter 1 reviews recent advances in enzyme-activated fluorogenic probes for biological imaging, organized by enzyme classification. This review surveys recent masking strategies, different modes of enzymatic activation, and the breadth of current and future probe applications. Key challenges, such as probe selectivity and spectroscopic requirements, are described in this chapter along with therapeutic and diagnostic opportunities that can be accessed by surmounting these challenges. Chapter 2. Electronic and Steric Optimization of Fluorogenic Probes for Biomolecular Imaging. In many fluorogenic probes, the intrinsic fluorescence of a small-molecule fluorophore is masked by ester masking groups until entry into a cell, where endogenous esterases catalyze the hydrolysis of esters, generating fluorescence. The susceptibility of masking groups to spontaneous hydrolysis is a major limitation of these probes. Previous attempts to address this problem have incorporated auto-immolative linkers at the cost of atom economy and synthetic adversity. In this chapter, I report on a linker-free strategy that employs adventitious electronic and steric interactions in easy-to-synthesize probes. I find that halogen-carbonyl n-->[pi]* interactions and acyl group size are optimized in 2',7'-dichlorofluorescein diisobutyrate. This probe is relatively stable to spontaneous hydrolysis but is a highly reactive substrate for esterases both in vitro and in cellulo, yielding a bright, photostable fluorogenic probe with utility in biomolecular imaging. Chapter 3. Cellular Uptake of Large Monofunctionalized Dextrans. Dextrans are a versatile class of polysaccharides with applications that span medicine, cell biology, food science, and consumer goods. In Chapter 3, I apply the electronically stabilized probe described in Chapter 2 to study the cellular uptake of a new type of large monofunctionalized dextran that exhibits unusual properties: efficient cytosolic and nuclear uptake. This dextran permeates various human cell types without the use of transfection agents, electroporation, or membrane perturbation. Cellular uptake occurs primarily through active transport via receptor-mediated processes. These monofunctionalized dextrans could serve as intracellular delivery platforms for drugs or other cargos. Chapter 4. Paired Nitroreductase-Probe System to Quantify the Cytosolic Delivery of Biomolecules. Cytosolic delivery of large biomolecules is a significant barrier to therapeutic applications of CRISPR, RNAi, and biologics such as proteins with anticancer properties. In Chapter 4, I describe a new paired enzyme-probe system to quantify cytosolic delivery of biomolecules-a valuable resource for elucidating mechanistic details and improving delivery of therapeutics. I designed and optimized a nitroreductase fusion protein that embeds in the cytosolic face of outer mitochondrial membranes, providing several key improvements over unanchored reporter enzymes. In parallel, I prepared and assessed a panel of nitroreductase-activated probes for favorable spectroscopic and enzymatic activation properties. Together, the nitroreductase fusion protein and fluorogenic probes provide a rapid, generalizable tool that is well-poised to quantify cytosolic delivery of biomolecules. Chapter 5. Future Directions. This chapter outlines several future directions for expanding the scope of fluorogenic probes and developing new biological applications. Additionally, Chapter 5 is followed by an appendix describing a tunable rhodol fluorophore scaffold for improved spectroscopic properties and versatility. Overall, the work described in this thesis illustrates the power of enzyme-activated fluorogenic probes to provide fresh insight into dynamic biological processes, with direct implications for improved therapeutic delivery. |
Quantifying and visualizing risk in the garment manufacturing supply chain | Supply chains are exposed to a variety of risks as they become more complex and geographically diverse. Disruptions due to these risks can be costly. Companies cannot hope to mitigate all of their supply chain risks. In order to focus risk management resources on locations in the supply chain with the most risk, companies need a comprehensive method to quantify all of their significant supply chain risks. We worked with a company in the garment manufacturing industry to map their supply chain for a few representative products. Using input from the company, we equated different risk indices with the probability of loss of a node in their supply chain. The probabilities of loss allowed us to calculate a value-at-risk at each node. Once calculated, the values-at-risk were overlaid on a visual depiction of the company's supply chain network. While previous studies have quantified and visualized risk in companies' supply chains, our research sought to combine different categories of risk in order to give a more comprehensive picture of the risk at each node. We looked at disruption risks due to natural disasters, supplier bankruptcy, and political instability. We found that commercially available indices that quantify different categories of risk can be used to inform supply chain risk management decisions. Moving from these indices to a value-at-risk model of a supply chain is not a wholly quantitative process. Therefore, the strength of the model lies more in the relative quantities of value-at-risk rather than their absolute values. Overlaying these values-at-risk over a visual depiction of their supply chain gave the company a clearer picture of where to focus risk management efforts. Other companies in other industries could apply a similar approach to build an organizational risk management tool. |
A study of electrowetting-assisted boiling | The classical theory of boiling heat transfer based on bubble dynamics is explained and includes a full derivation of the Rohsenow boiling correlation. An alternative, more accurate correlation for determining boiling heat transfer coefficients is then presented. The theory of DC and AC electrowetting is described along with electrolyis and electric breakdown, two common problems in successfully demonstrating electrowetting. Next, the construction of a 1000V DC power supply for electrowetting experiments is outlined along with failed and successful attempts to create surfaces that exhibit electrowetting. The design and construction of an electrowetting-compatible boiling apparatus capable of delivering up to 1000W of power to a 400mm2 sample is also detailed. The power supply, surfaces, and boiler are used to determine that while electrowetting increases surface wettability it also severely decreases heat transfer, disproving the initial hypothesis. Heat transfer per unit superheat, represented by the dimensionless ratio of the Nusselt number to the Jakob number, is found to increase as a double exponential with decay constants 0.23 ± 0.21 and -14.82 ± 49.69 as a function of the dimensionless electrowetting number, a measure of electrical energy to surface energy of a liquid. Likewise, thermal resistance as a function of the electrowetting number is found to increase as a double exponential with rise constants -37.95 ± 214.4 and -0.32 ± 2.23. |
Filming children of Agent Orange rehabilitation villages in Vietnam and the danger of misrepresentation | We represent the Vietnam War as a concluded event in the past; however, the Agent- Orange-affected population in Vietnam shows that war is contemporary. This population remains relatively unknown to the world - especially to the Western world. We are isolated from the Agent-Orange-affected population because we have grown dependent on curated images of the Vietnam War in popular media that do not include the local population. Here, a challenging duty of a filmmaker is to create new images that convey their experience. This thesis will examine the filmmaking process of the contemporary population affected by Agent Orange in Vietnam and raise larger questions about the ways in which we capture contemporary war victims' stories through video. How can film revive engagements with a seemingly concluded war? How do we tell stories of people with visually apparent abnormalities without ostracizing them? |
Kanban System design for automatic production scheduling | A Kanban Production System is designed to help a factory line meet fluctuating demands for multiple part types. Based on the parameter settings of the Control-Point Policy, the optimum Kanban levels are obtained. The simulation software Simul8 was used to model the factory line and the Kanban system. Using the optimum Kanban levels, the Kanban system will act as an automatic production scheduling system that will indicate clearly when and how much of each part-type should be produced. Use of this system will avoid unnecessary inventory and changeover cost incurred by the existing Kanban system used by the factory line. Key words: Kanban, Control-Point Policy, Optimization Disclaimer: The content of the thesis is modified to protect the real identity of the attachment company. Company name and confidential information are omitted. |
Inquiries into the behavior of emerging market firms | This dissertation addresses two aspects of firm behavior in emerging markets. Chapters 1 and 3 consider the decision of firms to borrow in domestic or foreign currency and the implications of this decision on fixed capital and inventory investment. Chapter 2 explores the effects of institutions, via transaction costs, on the variety of intermediate goods used by firms in different economies, and the effects of this choice on productivity levels. Much has been written recently about the problems for emerging markets that might result from a mismatch between foreign currency denominated liabilities and assets (or income flows) denominated in local currency. In particular, several models, developed in the aftermath of financial crises of the late 1990's, suggest that the expansion in the "peso" value of "dollar" liabilities resulting from a devaluation either aggravated or triggered many of the recent financial crises. However, little evidence has been presented either on the effects of foreign currency debt on investment or the variables that affect firm level debt choice in the first place. Two of the papers in this dissertation (Chapters 1 and 3) attempt to fill this gap using a new database with accounting information (including the currency composition of liabilities) for close to 400 non-financial firms in five Latin American countries. In Chapter 1 (coauthored with Hoyt Bleakley) we estimate, at the firm level, the reduced-form effect on investment of holding foreign currency denominated debt during an exchange rate realignment. We consistently find that this effect is positive, contrary to the predicted sign of the net-worth effect. |
Efficient migration path approach for deinterlacing 1080i high definition | Modern displays deinterlace interlaced high definition television (HDTV) to remove unsightly artifacts. U.S. over-the-air television broadcasting is scheduled to switch to an all-digital format in June, 2009. Therefore, we are ready for a new system which will improve HDTV. A receiver compatible system can add a small bitrate enhancement stream and enable new receivers to easily deinterlace l080i streams. Old receivers can ignore this stream while decoding the original lo8oi stream. Since the enhancement stream has a low bitrate, we can use this system alongside mobile digital television. The receiver compatible system shifts computing costs from the receivers to the transmitters. Since there are more receivers than transmitters, the net cost is reduced dramatically. The receiver compatible system has access to the original video, allowing for superior video quality compared to traditional deinterlacing systems. |
Evaluating style transfer in natural language | Style transfer is an active area of research growing in popularity in the Natural Language setting. The goal of this thesis is present a comprehensive review of style transfer tasks used to date, analyze these tasks, and delineate important properties and candidate tasks for future methods researchers. Several challenges still exist, including the difficulty of distinguishing between content and style in a sentence. While some state of the art models attempt to overcome this problem, even tasks as simple as sentiment transfer are still non-trivial. Problems of granularity, transferability, and distinguishability have yet to be solved. I provide a comprehensive analysis of the popular sentiment transfer task along with a number of metrics that highlight its shortcomings. Finally, I introduce possible new tasks for consideration, news outlet style transfer and non-parallel error correction, and provide similar analysis for the feasibility of using these tasks as style transfer baselines. |
High-throughput quantification of glycoprotein sialylation | Sialic acid can improve qualities of therapeutic glycoproteins, such as circulatory half-life, biological activity, and solubility. In production of therapeutic glycoproteins, a high-throughput method (HTM) is required for process monitoring and optimization to ensure consistent and optimal sialic acid content. The HTM is also required for cell clone screening in cell line development. Current methods for quantifying sialic acid, however, require chromatographic separation that is time consuming and cannot rapidly analyze many samples in parallel. Here we develop a novel HTM for quantifying glycoprotein sialylation. Using chemical reduction, enzymatic release of sialic acid, and chemical derivatization, the HTM can accurately, rapidly (15 min), and specifically analyze many samples in parallel. It requires only 45 piL of sample and has a quantitation limit of 2 ptM sialic acid. We validated the HTM for monitoring sialylation of recombinant interferon-gamma (IFN-y) produced in Chinese Hamster Ovary (CHO) cell culture. The HTM was accurate in monitoring sialylation of IFN-y in batch CHO cell cultures. Furthermore, we used the HTM to study the effects of feeding ManNAc, Cu 2+, and Mn2+ on sialylation of glycoproteins produced in CHO-IFN-y cell cultures. We found that feeding these chemicals increased sialylation from 20 to 36 mg sialic acid/g protein in batch CHO cell cultures. Moreover, a quadratic least square model predicts that the feeding 2 mM ManNAc and 100 pM Cu2+ will increase the sialylation to 41 +/- 4 mg sialic acid/g protein, close to the experimental value of 35 ± 5 mg sialic acid/g protein. We also used the HTM to study intraclonal variability in glycoprotein sialylation. We found that there was significant variability in sialic acid content and productivity. The sialic acid content varied from 1 to 70 mg sialic acid/g protein, and was negatively correlated with the productivity. Overall, we have developed a novel HTM and demonstrated its versatility for various applications in bioprocesses. The HTM can measure sialic acid content of hundreds of samples in 15 minutes, while conventional methods require more than one day per sample. Thus, the HTM is an important analytical tool for producing therapeutic proteins with consistent and optimum sialylation. |
Responses from electric stimulation of CN | Cochlear nucleus (CN), the exclusive destination of the auditory nerve, is the gateway for all central processing of auditory information. The CN comprises three major subdivisions: anteroventral, posteroventral and dorsal (AVCN, PVCN and DCN, respectively), each of which contains anatomically and physiologically distinct neurons projecting onto different targets. This research used focal electric stimulation of small, confined parts of the CN in anesthetized guinea pigs to resolve the roles of the CN divisions, in two contexts. Part i explored the effect of stimulation on the gross neural potential (electrically evoked auditory brainstem response, EABR). In AVCN and PVCN away from the 8th nerve fibers entering the brainstem, stimulation consistently evoked waveforms comprising 3 waves, suggesting a diffuse distribution of cellular generator of the EABR. On the other hand, in vestibular structures (vestibular nerve root and Scarpa's ganglion), the characteristic waveform comprised only two waves. Stimulation of multiple neural structures, as seen with higher stimulus levels or stimulation in auditory nerve root area generally produced more complex and variable waveforms. Part 2 explored the effects of stimulation on the activation of one type of auditory reflex, medial olivocochlear (MOC) reflex. The reflex was monitored through its effects on distortion product otoacoustic emission (DPOAE). The MOC reflex was activated bilaterally by stimulating PVCN or AVCN shell, but not AVCN core. These results suggest that there are two groups of MOC interneurons in specific parts of CN. |
Scale effects in microindentation of ductile crystals | Indentation testing has long been a standard test used to classify all types of materials. In the past several decades the scale of indentation testing has moved into the micron and even sub-micron range. For many types of materials, at these small length scales, the hardness of the material measured by the indentation test depends on the depth of the indentation. This indentation size effect was not observed at the larger length scales. Because indentation testing with conical or pyramidal indenter tips is geometrically similar, the existence of a size effect was surprising. Since the size effect associated with microindentation was discovered, many theories about its cause have been proposed. Several of the theories suggest that the source of the indentation size effect is experimental error. Such factors as inaccurate measurement of the contact area, indenter tip deformities, improper surface preparation, lateral movement of the indenter tip, inhomogeneity of the material, compliance of the test fixture, anisotropic deformations, thermal drift, and noise have been cited as areas where experimental error may play a role in the size effect. Another group of theories suggests that there are actual physical causes for the size effect. Some of the proposed physical causes of the size effect are friction between the specimen and the indenter tip, elastic recovery of the indent, material pile-up and sink- in, work-hardened surface material, oxidized surface layers, and variation of material parameters due to the stress state of the material. One area that has received some attention recently as a possible cause of the indentation size effect is hardening resulting from geometrically necessary dislocations (GNDs). GNDs arise due to curvature in the crystalline lattice from gradients of plastic shear strain. As the indentation depth decreases, the relative strain gradients within the test specimen increase. |
The use of satellite remote sensing to determine land water runoff | The purpose of this study is to estimate land water runoff from satellite measurements of net radiation at the top of the atmosphere by invoking columnar atmospheric water and energy balance. Land water runoff is an important parameter in many areas including management of human water use and validation of climate models. Energy measurements taken by satellite are less prone to error than the river discharge measurements required in conventional land water runoff estimates. Mean annual )and water runoff can be determined from net radiation at the top of the atmosphere and the Divergence Bowen Ratio (DBR). The DBR is the ratio of the horizontal flux of internal and potential energy to the horizontal flux of latent heat in the atmosphere. A tropical DBR of -2 was determined from two independent techniques, from a least squares regression relating net radiation with runoff and from the calculation of DBR from local surface humidity. The DBR of -2 corresponds to an approximately one-to-one relationship between net radiation at the top of the atmosphere and land water runoff in the tropics. This DBR also implies that the horizontal divergence of internal and potential energy is equal to twice the horizontal convergence of latent heat through a vertical atmospheric column in the tropics. These simple relationships do not hold in the higher latitudes. Over the tropics, horizontal and vertical transports of energy and moisture are predominantly by nearsteady air motions. Across middle to high latitudes, transient eddy mechanisms become important and the assumptions of the DBR (horizontal temperature homogeneity and steady circulation) break down. |