Engineering Commons^™

An Efficient Privacy-Preserving Framework For Video Analytics, Tian Zhou

Doctoral Dissertations

With the proliferation of video content from surveillance cameras, social media, and live streaming services, the need for efficient video analytics has grown immensely. In recent years, machine learning based computer vision algorithms have shown great success in various video analytic tasks. Specifically, neural network models have dominated in visual tasks such as image and video classification, object recognition, object detection, and object tracking. However, compared with classic computer vision algorithms, machine learning based methods are usually much more compute-intensive. Powerful servers are required by many state-of-the-art machine learning models. With the development of cloud computing infrastructures, people are able …

Incorporating Machine Learning With Satellite Data To Support Critical Infrastructure Measurement And Sustainable Development, Aggrey Muhebwa

Doctoral Dissertations

Under the umbrella concept of Artificial Intelligence (AI) for good, recent advances in machine learning and large-scale data analysis have opened new opportunities to solve humanity’s most pressing challenges. Improvements in computation complexity and advances in AI (e.g., Vision Transformers) have led to faster and more effective techniques for extracting high-dimensional patterns from large-scale heterogeneous datasets (big data). Further, as satellite data become increasingly available at varying temporal-spatial resolutions, AI tools are helping us to better understand the underlying causes of environmental and socioeconomic changes at an unprecedented scale, ushering in an era of data-driven decision-making to support sustainable and …

Machine Learning Modeling Of Polymer Coating Formulations: Benchmark Of Feature Representation Schemes, Nelson I. Evbarunegbe

Masters Theses

Polymer coatings offer a wide range of benefits across various industries, playing a crucial role in product protection and extension of shelf life. However, formulating them can be a non-trivial task given the multitude of variables and factors involved in the production process, rendering it a complex, high-dimensional problem. To tackle this problem, machine learning (ML) has emerged as a promising tool, showing considerable potential in enhancing various polymer and chemistry-based applications, particularly those dealing with high dimensional complexities.

Our research aims to develop a physics-guided ML approach to facilitate the formulations of polymer coatings. As the first step, this …

Measuring Accessibility To Food Services To Improve Public Health, Efthymia Kostopoulou

Masters Theses

Food accessibility has lately been of primary interest given its impact on public health outcomes. This thesis illustrates the gaps in food access by applying spatial analysis in Massachusetts accounting for a variety of demographic and socioeconomic factors. The number of grocery stores, farmers markets, and convenience stores within 1/4 and 1 mile of the Census tracts’ centroids are the two accessibility metrics used in the spatial analysis. In addition, a regression model is developed using the Gradient Boosting machine learning method to show the relationship between the socioeconomic factors and the number of grocery stores within 1 mile of …

Enhancing Management Of Built And Natural Water And Sanitation Systems With Data Science, Nelson Da Luz

Doctoral Dissertations

In the age of the data revolution, the civil engineer can enhance the management of infrastructure systems using new techniques focused on data. This dissertation present three studies in which data science approaches are used to enhance management of water and sanitation systems in both the built and natural environments. Chapters 1 and 2 focus on improving methods for data collection relating to water quality monitoring. In Chapter 1, the efficacy of different water quality sampling program designs is evaluated as the programs relate to meeting monitoring goals. Considerations include how timing, location, and distribution system operations can affect monitoring …

Numerical Modeling Of Advanced Propulsion Systems, Peetak P. Mitra

Doctoral Dissertations

Numerical modeling of advanced propulsion systems such as the Internal Combustion Engine (ICE) is of great interest to the community due to the magnitude of compute/algorithmic challenges. Fuel spray atomization, which determines the rate of fuel-air mixing, is a critical limiting process for the phenomena of combustion within ICEs. Fuel spray atomization has proven to be a formidable challenge for the state-of-the-art numerical models due to its highly transient, multi-scale, and multi-phase nature. Current models for primary atomization employ a high degree of empiricism in the form of model constants. This level of empiricism often reduces the art of predictive …

Short Term Energy Forecasting For A Microgird Load Using Lstm Rnn, Akhil Soman

Masters Theses

Decentralization of the electric grid can increase resiliency (during natural disasters) and can reduce T&D energy losses and emissions. Microgrids and DERs can enable this to happen. It is important to optimally control microgrids and DERs to extract the greatest economic, environmental and resiliency benefits. This is enabled by robust forecasting to optimally control loads and energy sources. An integral part of microgrid control is power side and load side demand forecasting.

In this thesis, we look at the ability of a powerful neural network algorithm to forecast the load side demand for a microgrid using the UMass campus as …

Computational Modeling Of The Structure And Catalytic Behavior Of Graphene-Supported Pt And Ptru Nanoparticles, Raymond Gasper

Doctoral Dissertations

Computer modeling has the potential to revolutionize the search for new catalysts for specific applications primarily via high-throughput methodologies that allow researchers to scan through thousands or millions of potential catalysts in search of an optimal candidate. To date, the bulk of the literature on computational studies of heterogeneous catalysis has focused on idealized systems with near-perfect crystalline surfaces that are representative of macroscopic catalysts. Advancing the frontier to nanoscale catalysis, in particular, heterogeneous catalysis on nanoclusters, requires consideration of low-symmetry nanoparticles with realistic structures including the attendant complexity arising from under-coordination of catalyst atoms and dynamic fluxionality of clusters. …

Aeroelastic Simulation Of Wind Turbines Using Free Vortex Methods And Strategies For Accelerating The Computation, Shujian Liu

Doctoral Dissertations

This dissertation integrated the free vortex method code Wake Induced Dynamics Simulator (WInDS), which was developed by Sebastian et al., into the open source and widely-used software FAST. A range of computational strategies including paral- lelization and Treecode algorithms are used to increase the computational efficiency of the software. Full aero-hydro-servo-elastic simulations with free vortex method are conducted, which focus on an in-depth study on the influence of the aeroelasticity of the wind turbine and platform motions on the unsteadiness of the aerodynamics, and the comparison of aeroelastic responses of two floating wind turbine concepts. This dissertation also applies long …

Adaft: A Resource-Efficient Framework For Adaptive Fault-Tolerance In Cyber-Physical Systems, Ye Xu

Doctoral Dissertations

Cyber-physical systems frequently have to use massive redundancy to meet application requirements for high reliability. While such redundancy is required, it can be activated adaptively, based on the current state of the controlled plant. Most of the time the physical plant is in a state that allows for a lower level of fault-tolerance. Avoiding the continuous deployment of massive fault-tolerance will greatly reduce the workload of CPSs. In this dissertation, we demonstrate a software simulation framework (AdaFT) that can automatically generate the sub-spaces within which our adaptive fault-tolerance can be applied. We also show the theoretical benefits of AdaFT, and …

Explorations Into Machine Learning Techniques For Precipitation Nowcasting, Aditya Nagarajan

Masters Theses

Recent advances in cloud-based big-data technologies now makes data driven solutions feasible for increasing numbers of scientific computing applications. One such data driven solution approach is machine learning where patterns in large data sets are brought to the surface by finding complex mathematical relationships within the data. Nowcasting or short-term prediction of rainfall in a given region is an important problem in meteorology. In this thesis we explore the nowcasting problem through a data driven approach by formulating it as a machine learning problem.

State-of-the-art nowcasting systems today are based on numerical models which describe the physical processes leading to …

Intrinsic Functions For Securing Cmos Computation: Variability, Modeling And Noise Sensitivity, Xiaolin Xu

Doctoral Dissertations

A basic premise behind modern secure computation is the demand for lightweight cryptographic primitives, like identifier or key generator. From a circuit perspective, the development of cryptographic modules has also been driven by the aggressive scalability of complementary metal-oxide-semiconductor (CMOS) technology. While advancing into nano-meter regime, one significant characteristic of today's CMOS design is the random nature of process variability, which limits the nominal circuit design. With the continuous scaling of CMOS technology, instead of mitigating the physical variability, leveraging such properties becomes a promising way. One of the famous products adhering to this double-edged sword philosophy is the Physically …

On Physical Disorder Based Hardware Security Primitives, Arunkumar Vijayakumar

Doctoral Dissertations

With CMOS scaling extending transistors to nanometer regime, process variations from manufacturing impacts modern IC design. Fortunately, such variations have enabled an emerging hardware security primitive - Physically Unclonable Function. Physically Unclonable Functions (PUFs) are hardware primitives which utilize disorder from manufacturing variations for their core functionality. In contrast to insecure non-volatile key based roots-of-trust, PUFs promise a favorable feature - no attacker, not even the PUF manufacturer can clone the disorder and any attempt at invasive attack will upset that disorder. Despite a decade of research, certain practical problems impede the widespread adoption of PUFs. This dissertation addresses the …