As the use of unmanned vehicles continues to grow, so does the need for systems to safely test and operate these vehicles. While there are safety systems designed for this purpose, they are often developed for a specific vehicle platform. The Multi-Level Emergency Stop (MLES) system provides three user-defined emergency response contingencies that can be adapted to a wide variety of unmanned vehicles. The Multi-Level Emergency Stop system is designed to be an ad-on safety system that can be integrated into ground, air, or surface unmanned vehicles. A complete MLES system consists of a hand held transmitter and a vehicle mounted receiver. The three levels of contingencies are controlled by three switches on the transmitter. These switches engage and disengage contacts located in the receiver via a wireless link. The function of these contacts is determined by the user for each unique application. Presented in this thesis is the detailed hardware design and software layout of the Multi-Level Emergency Stop system. Also included are the performance results and operational tests. / Master of Science

We consider a multi-task learning problem, where an agent is presented a number of N reinforcement learning tasks. To solve this problem, we are interested in studying the gradient approach, which iteratively updates an estimate of the optimal policy using the gradients of the value functions. The classic policy gradient method, however, may be expensive to implement in the multi-task settings as it requires access to the gradients of all the tasks at every iteration. To circumvent this issue, in this paper we propose to study an incremental policy gradient method, where the agent only uses the gradient of only one task at each iteration. Our main contribution is to provide theoretical results to characterize the performance of the proposed method. In particular, we show that incremental policy gradient methods converge to the optimal value of the multi-task reinforcement learning objectives at a sublinear rate O(1/√k), where k is the number of iterations. To illustrate its performance, we apply the proposed method to solve a simple multi-task variant of GridWorld problems, where an agent seeks to find an policy to navigate effectively in different environments. / Master of Science / First, we introduce a popular machine learning technique called Reinforcement Learning (RL), where an agent, such as a robot, uses a policy to choose an action, like moving forward, based on observations from sensors like cameras. The agent receives a reward that helps judge if the policy is good or bad. The objective of the agent is to find a policy that maximizes the cumulative reward it receives by repeating the above process. RL has many applications, including Cruise autonomous cars, Google industry automation, training ChatGPT language models, and Walmart inventory management. However, RL suffers from task sensitivity and requires a lot of training data. For example, if the task changes slightly, the agent needs to train the policy from the beginning. This motivates the technique called Multi-Task Reinforcement Learning (MTRL), where different tasks give different rewards and the agent maximizes the sum of cumulative rewards of all the tasks. We focus on the incremental setting where the agent can only access the tasks one by one randomly. In this case, we only need one agent and it is not required to know which task it is performing. We show that the incremental policy gradient methods we proposed converge to the optimal value of the MTRL objectives at a sublinear rate O(1/ √ k), where k is the number of iterations. To illustrate its performance, we apply the proposed method to solve a simple multi-task variant of GridWorld problems, where an agent seeks to find an policy to navigate effectively in different environments.

The purpose of this study is to examine the relationship among student civic knowledge scores and several different variables each at the student, classroom/school, and national levels using the IEA CIVED study international data set collected in 1999 from 27 countries. The student level predictors included two elements of socioeconomic status (a student's parental education, their home literacy level measured by the number of books at home), student's perception of an open classroom climate, student aspiration of obtaining higher education, and other variables that were identified as relevant to the dependent variable in the literature. The classroom/school level predictors included teacher's degree in civics, in-service training, teaching confidence, and school safety in addition to the compositional variable created as the classroom/school averages by aggregating the student level variables. Then I investigated whether instructional methods focusing on the student activities the teacher employed in the classroom and an open classroom climate were associated after accounting for the above student and school level background variables. National level variables such as GNP, GINI index, democratic system, public education expenditure, and etc. as well as compositional variables obtained by aggregating the classroom/school variables were also added to the model to investigate if they were associated with students' civic knowledge scores and whether they could explain between nations variability. The study used a three-level hierarchical linear model to analyze the data, with number of students, N=56,579, number of classrooms/schools, J=3443, and number of countries, K=27. Some of the key findings was that there were significant variations of civics knowledge among nations, and significant variations of civic knowledge scores between school and within nations, no statistically significant association between teacher's practice and civics knowledge scores, however the student perception of an open classroom climate was significant at all 3 levels. These findings were interpreted in terms of policies and practices that could be implemented to improve students' civic knowledge. / Ph. D.

The limited available budget along with old aging infrastructure in nation magnifies the role of strategic decision making for maintenance of infrastructure. The challenging objective is to maintain the infrastructure asset systems in a state of good repair and to improve the efficiency and performance of the infrastructure systems while protecting and enhancing the natural environment. Decision makers are in need of a decision support system to consider these multiple objectives and criteria to effectively allocate funding and achieve the highest possible return on investment on their infrastructure. The research proposes and validates a framework for such decisions. The proposed model aims at finding optimal techniques for maintenance of multiple roadway asset items while taking into account time, cost, level of service and environmental impacts. Therefore, the goal is to answer what are the optimal combinations of maintenance techniques for roadway assets while more than one objective is being optimized. In other words, the main objective is to develop a decision support system for selecting and prioritizing necessary actions for MRandR (Maintenance, Repair and Rehabilitation) of multiple asset items in order for a roadway to function within an acceptable level of service, budget, and time while considering environmental impacts. To achieve these desirable outcomes, this model creates a two-stage framework for a sustainable infrastructure asset management. First a multi-objective problem based on the multi colony ant colony optimization is analyzed. The objectives of the problem are: (i) Minimizing maintenance costs, (ii) Minimizing maintenance time, (iii) Minimizing environmental impacts and (iv) Maximizing level of service improvement. In the second stage, the results of the multi objective optimization will be prioritized using a Multi Criteria Decision Making (MCDM) process. The proposed approach will simultaneously optimize four conflicting objectives along with using a multi criteria decision-making technique for ranking the resulted non-dominated solutions of multi objective optimization. The results of implementation of the proposed model on a section of I-64 highway are presented for a sub-set of asset items. Moreover, the proposed model is validated using a scalable test problem as well as comparison with existing examples. Results reveal the capability of the model in generation of optimal solutions for the selection of maintenance strategies. The model optimizes decision making process and benefits decision makers by providing them with solutions for infrastructure asset management while meeting national goals towards sustainability and performance-based approach. In addition, provides a tool to run sensitivity analysis to evaluate annual budget effects and environmental impacts of different resource allocation scenarios. Application of the proposed approach is implemented on roadway asset items but it is not limited to roadways and is applicable to other infrastructure assets. / Ph. D.

When experiencing a building's interior or exterior conditions, one may be inclined to 'feel-out' its spatial and volumetric proportions, judge their appropriateness, its quality of formal conditions, its power, its clearness of the structure, and get a sense for the way its architecture was placed onto the site. It is said that, 'knowledge is key', and knowing how a building is soundly and structurally assembled and seated onto the earth -- is key. This thought brings to the table an important question, why do we build beautifully sound and monolithic (at times) structural systems then choose to cover them up entirely? In the context of the Washington, D.C.'s current building climate -- why must we build a dense grove of slender wood posts atop concrete plinths only to cover them up in clothing with certain ephemeral stylistic ideas? Obvious reasons such as insulation and weatherproofing are valid, but thermal barrier technology now allows for exposing the raw architectural elements without sacrificing thermal qualities. Can we use this technology to our advantage, and if so, how would one begin to conceive of a structural system which celebrates the bearing members in an architectural manner? Are there ways to interact more directly with the structure itself? In what manner will the site specific and environmental constraints play a role in making creative architectural decisions? I believe the research conducted in the past year resulted in a truthful approach toward form finding, space making, and respecting the chosen site and its unique constraints. / Master of Architecture

Distributed sensing has been an attractive alternative to the traditional single-point sensing technology when measurement at multiple locations is required. Traditional distributed sensing methods based on silica optical fiber and electric coaxial cables have some limitations for specific applications, such as in smart textiles and wearable sensors. By adopting the fiber thermal drawing technique, we have designed and fabricated multi-material electrode-embedded polymer fibers with distributed sensing capabilities. Polymers sensitive to temperature and pressure have been incorporated into the fiber structure, and thin metal electrodes placed inside fiber by convergence drawing have enabled detection of local impedance change with electrical reflectometry. We have demonstrated that these fibers can detect temperature and pressure change with high spatial resolution. We have also explored the possibility of using polymer optical fiber in a Raman scattering based distributed temperature sensing system. Stokes and Anti-Stokes signals of a PMMA fiber illuminated by a 532 nm pulsed laser was recorded, and the ratio was used to indicate local temperature change. We have also developed a unique way to fabricate porous polymer by thermal drawing polymer materials with controlled water content in the polymer. The porous fibers were loaded with a fluorescent dye, and its release in tissue phantoms and murine tumors was observed. The work has broadened the scope of multi-material, multi-functional fiber and may shed light on the development of novel smart textile devices. / PHD / In recent years smart textiles and wearable gadgets have already changed the way we live. There has been increasing industrial interest to develop novel flexible, stretchable devices that can interact with human and the environment. Thermal drawing technique originally invented for manufacturing telecommunication optical fiber has been used by researchers to fabricate fibers with more functionality. In this work, we report the progress made on the fabrication of multi-material fiber. Soft polymer fibers capable of measuring temperature and pressure were designed and made by the thermal drawing technique. Submillimeter fibers with thin copper electrodes have shown potential to be readily embedded in a smart fabric to provide 1D information in one direction or woven into a 2D pattern for area monitoring. We have also explored another temperature measurement scheme using polymer optical fibers with a pulsed laser. Compared with the electronic fibers, it is less susceptible to electrical noise and more robust. Lastly, we have shown a unique way to generate porosity in thermally drawn polymer fibers. The elongated pores in the fibers come from water escaping the fiber during the fabrication process. The three aspects of the project expand the scope of multi-material, multi-functional fiber and can shed light on the future development of electronic textile devices.

Axial ventilation fans are used to improve the air quality, remove contaminants, and to control the temperature and humidity in occupied areas. Ventilation fans are one of the most harmful sources of noise due to their close proximity to occupied areas and widespread use. The prolonged exposure to hazardous noise levels can lead to noise-induced hearing loss. Consequently, there is a critical need to reduce noise levels from ventilation fans. Since fan noise scales with the 4-6th power of the fan tip speed, minimizing the fan tip speed and optimizing the duct geometry are effective methods to reduce fan noise. However, there is a tradeoff between reducing fan speed, noise and aerodynamic efficiency. To this end, a new innovative comprehensive optimum design methodology considering both aerodynamic efficiency and noise was formulated and implemented using a multi-objective genetic algorithm. The methodology incorporates a control vortex design approach that results in a spanwise chord and twist distribution of the blades that maximize the volumetric flow rate contribution of the outer radii, i.e. the axial flow velocity increases from the fan hub to the tip. The resulting blade geometry generates a given volumetric flow rate at the minimum fan tip speed. The fan design is complemented by the design of the optimum inlet duct geometry to maximize volumetric flow rate and minimize BL thickness for low noise generation. Good agreement with experimental results validates the design process. The present study also incorporates multi-element airfoils to further increase the aerodynamic characteristics of the fan blades and enable lower fan speeds and noise. Good agreement between experiments and predictions indicate that traditional blade element momentum methods can be implemented in conjunction with multi-element airfoil aerodynamic characteristics with good accuracy. A direct comparison of fans designed with single and multi-element airfoils has shown that fans designed with multi-element airfoils aerodynamically outperform single element fans. / Doctor of Philosophy / Axial ventilation fans are widely used to improve the air quality, remove contaminants, and to control the temperature and humidity in occupied areas. However, high noise levels from ventilations fans are a harmful source of noise that can lead to irreversible noise-induced hearing loss. Therefore, this work addresses a critical need for quiet and efficient ventilation fans. To this end, a new innovative comprehensive optimum design methodology considering both aerodynamic efficiency and noise was formulated, implemented, and tested. The methodology optimizes the fan geometry to maximize the volumetric flow rate and minimize noise. The fan design is complemented by the design of the optimum inlet duct geometry to increase the volumetric flow rate and minimize BL thickness for low noise generation. Good agreement with experimental results validates the design process. The present study also incorporates multi-element airfoils to further increase the aerodynamic characteristics of the fan blades. A direct comparison of fans designed with single and multi-element airfoils has shown that fans designed with multi-element airfoils aerodynamically outperform single element airfoil fans.

This work investigates the use of multi-degree of freedom (MDOF) passive and active vibration absorbers for the control of structural vibration as an improvement to conventional single degree of freedom (SDOF) vibration absorbers. An analytical model is first used to compare passive two degree of freedom (2DOF) absorbers to SDOF absorbers using point impedance as the performance criterion. The results show that one 2DOF absorber can provide the same impedance at two resonance frequencies as two SDOF absorbers for equal amounts of total mass. Experimental testing on a composite cylindrical shell supports the assertion that a 2DOF absorber can attenuate two resonance frequencies. Further modeling shows that MDOF absorbers can utilize the multiple mode shapes that correspond to their multiple resonance frequencies to couple into modes of a distributed primary system to improve the attenuation of structural resonance. By choosing the coupling positions of the MDOF absorber such that its mode shape mirrors that of the primary system, the mass of the absorber can be utilized at multiple resonance frequencies. For limited ranges of targeted resonance frequencies, this technique can result in MDOF absorbers providing attenuation equivalent to SDOF absorbers while using less mass. The advantage gained with the MDOF absorbers is dependent on the primary system. This work compares the advantage gained using the MDOF absorbers for three primary systems: MDOF lumped parameter systems, a pinned-pinned plate, and a cylindrical shell. The active vibration absorber study in this work is highly motivated by the desire to reduce structural vibration in a rocket payload fairing. Since the efficiency of acoustic foam is very poor at low frequencies, the target bandwidth was 50 to 200 Hz. A 2DOF active vibration absorber was desired to exhibit broad resonance characteristics over this frequency band. An analytical model was developed to facilitate the design of the mechanical and electrical properties of the 2DOF active vibration absorber, and is supported by experimental data. Eight active vibration absorbers were then constructed and used in a multiple-input multiple-output (MIMO) feed-forward control system on a mock payload fairing under high level acoustic excitation. The results show significant levels of global attenuation within the targeted frequency band. / Master of Science

With the advances in VLSI and System-On-Chip (SOC) technologies, the complexity of hardware systems has increased manifold. The increasing complexity poses serious challenges to the digital hardware design. Functional verification has become one of the most expensive and time-consuming components of the current product development cycle. Today, design verification alone often surpasses 70% of the total development cost and the situation has been projected to continue to worsen. The two most widely used formal methods for design verification are Equivalence Checking and Model Checking. During the design phase, hardware goes through several stages of optimizations for area, speed, power, etc. Determining the functional correctness of the design after each optimization step by means of exhaustive simulation can be prohibitively expensive. An alternative to prove functional correctness of the optimized design is to determine the design's functional equivalence with respect to some golden model which is known to be functionally correct. Efficient techniques to perform this process is known as Equivalence Checking. Equivalence Checking requires that the implementation circuit should be functionally equivalent to the specification circuit. Complexities in Equivalence Checking can be exponential to the circuit size in the worst case. Since Equivalence Checking of sequential circuits still remains a challenging problem, in this thesis, we first address this problem using efficient learning techniques. In contrast to the traditional learning methods, our method employs a mining algorithm to discover global constraints among several nodes efficiently in a sequential circuit. In a Boolean satisfiability (SAT) based framework for the bounded sequential equivalence checking, by taking advantage of the repeated search space, our mining algorithm is only performed on a small window size of unrolled circuit, and the mined relations could be reused subsequently. These powerful relations, when added as new constraint clauses to the original formula, help to significantly increase the deductive power for the SAT engine, thereby pruning a larger portion of the search space. Likewise, the memory required and time taken to solve these problems are alleviated. We also propose a pseudo-functional test generation method based on effective functional constraints extraction. We use mining techniques to extract a set of multi-node functional constraints which consists of illegal states and internal signal correlation. Then the functional constraints are imposed to a ATPG tool to generate pseudo functional delay tests. / Master of Science

Determining the optimum computer architecture configuration for a specific application or a generic algorithm is a difficult task. The complexity involved in today's computer architectures and systems makes it more difficult and expensive to easily and economically implement and test full functional prototypes of computer architectures. High level VHDL performance modeling of architectures is an efficient way to rapidly prototype and evaluate computer architectures. Determining the architecture configuration is fixed, one would like to know the tolerance and expected performance of individual/critical components and also what would be the best way to map the software tasks onto the processor(s). Trade-offs and engineering compromises can be analyzed and the effects of certain component failures and communication bottle-necks can be studied. A part of the research work done for the RASSP (Rapid Prototyping of Application Specific Signal Processors) project funded by Department of Defense contracts is documented in this thesis. The architectures modeled include a single-processor, single-global-bus system; a four processor, single-global-bus system; a four processor, multiple-local-bus, single-global-bus system; and finally, a four processor multiple-local-bus system interconnected by a crossbar interconnection switch. The hardware models used are mostly legacy/inherited models from an earlier project and they were upgraded, modified and customized to suit the current research needs and requirements. The software tasks that are run on the processors are pieces of the signal and image processing algorithm run on the Synthetic Aperture Radar (SAR). The communication between components/devices is achieved in the form of tokens which are record structures. The output is a trace file which tracks the passage of the tokens through various components of the architecture. The output trace file is post-processed to obtain activity plots and latency plots for individual components of the architecture. / Master of Science