Project Portfolio

The "Computer Vision Based Beacon Localization" project aims to determine the position of a beacon relative to an optical sensor. Employing advanced technologies like OpenCV and TensorFlow, this innovative endeavor navigates the challenge of interfacing multiple stages of deep learning technologies to work cohesively. Although the project is currently a work-in-progress, it already showcases significant strides in the application of computer vision and deep learning to real-world localization problems. 

The "Molten Borate Salt Rheological Studies" project is a focused effort to characterize the density and viscosity of molten borate salt under different gas environments and temperatures. Confronting inherent challenges such as the rapid cooling of the salt and small sample environments, a concentric cylinder rheometer was utilized. This determined endeavor led to first-of-its-kind measurements, successfully characterizing the salt's density and viscosity at temperatures up to 850°C.

The "Molten Salt Research Reactor Gas Dynamics Analysis" project was undertaken to predict the influence of circulating gas voids on the system nuclear behavior of the reactor, given the dual-phase nature of the molten salt used. This involved facing unprecedented challenges requiring large-scale high-performance computing, with the necessity of accurately determining the density of data points in parameter space. This challenge was overcome by initially conducting minor scoping studies before committing to full-scale studies. The SCALE reactor physics code was used throughout the process, and an innovative theory of gas behavior in Molten Salt Reactors (MSRs) was developed as part of the project. The analysis concluded successfully, making a significant contribution to the preliminary safety assessment report (PSAR) submitted to the Nuclear Regulatory Commission (NRC).

The "ARPA-E Molten Salt Pump Analysis and Development" project involved a critical contribution to understanding the underlying physics and limitations of the canned rotor pump. The goal was to discern the restricting factors when transitioning the design from kilo-watt to mega-watt levels. This task necessitated a deep comprehension of materials, electromechanical machine design, mechanical design, and stress analysis. A standout element of this project was the development process of a trailblazing pump design, aiming to be the first of its kind with an actively magnetically levitated bearing, operational at temperatures beyond 700°C. The project culminated with the development of a predictive code, capable of determining the magnitude of key design outputs based on various design parameters.

The "Molten Salt Reactor Experiment Xenon Analysis" project was an endeavor to model the effects of xenon in the MSRE. This task required navigating through decades of poorly-conditioned and often contradicting MSRE documentation, necessitating a deep knowledge of the MSRE program. A unique aspect of this project was the development of a code in Simulink, the first of its kind to account for the effects of salt and gas recirculation in the pump bowl. As a result of this intensive work, the project successfully developed the first MSR xenon analysis code that reached a substantial level of validation.

The "Gamma Camera Modeling and Simulation" project involved the creation of a ray-tracing code for parametric gamma camera designs, addressing the limitations of Monte Carlo methods due to restricted computational resources. This challenge was overcome by developing a C# library for gamma ray tracing and attenuation modeling in three dimensions. This novel code filled a gap between simpler programs such as Microshield and more complex ones like MCNP. The successful outcome was a unique tool that enabled a comprehensive parametric analysis of various gamma camera designs.

The "Mechatronic Positioning System" was a meticulously crafted project focusing on the design, development, and construction of a modular, mobile positioning system capable of accurate actuator placement within 1mm precision across a 3m x 3m area. A key application involved the positioning of an autonomous rail-based carriage relative to a target using LIDAR technology. The journey of this project was comprehensive, spanning from initial conceptual design through detailed design, procurement, construction, and finally, testing and debugging of the system. A significant challenge of managing slip in the friction drive was effectively tackled using a tensioning spring and integrating full feedback into the control algorithm. This project uniquely transitioned from concept to a fully operational prototype in under four months, leveraging friction drive and LIDAR technologies. The project's culmination was marked by the successful production and comprehensive testing of the system, showcasing its remarkable precision and the promise of mechatronics and LIDAR technology in positioning systems.

The "Plasma Chamber and Neutron Generator" project involved the design, development, and construction of a plasma chamber utilizing inertial electrostatic confinement to generate plasma. When loaded with deuterium, the chamber could potentially facilitate nuclear fusion and produce 2.45 MeV neutrons through the DD reaction. The main challenge was sourcing high vacuum equipment due to the scarcity of suppliers in Canada. This was overcome, enabling the use of inertial electrostatic confinement (IEC), high vacuum, two stages of vacuum gauges (Pirani and ionization), and a Cockcroft-Walton voltage multiplier. This project marked the first attempt at creating a "fusor" type device in Canada, with the successful construction and testing of the plasma chamber serving as the final result.

The "Optical Tomography Lab" project was an innovative endeavor designed to provide remote laboratory exercises for a radiological health program. The project revolved around developing exercises involving a CT machine, detailed image processing, and the implementation of "contrast-enhanced subtraction." A significant challenge was the presence of undocumented data formats that necessitated extensive reverse engineering. To tackle this, a blend of MATLAB and C# was used to develop algorithms, and a secure remote access system was built utilizing RDP. This pioneering project allowed students to conduct laboratory exercises entirely remotely with minimal supervision, representing a significant stride in flexible learning. The project concluded successfully, with the lab exercises fully developed and ready for student instruction.

The "AmBe Neutron Source Modeling" project was centered around the detailed modeling of the Americium Beryllium neutron source at Ontario Technical University using MCNP for precise dosimetry and safety calculations. Navigating through limited data availability for the AmBe Neutron spectrum, the project managed to utilize relevant data published by the IAEA to achieve its goals. This work marked the first time the neutron source had been modeled and was undertaken shortly after commissioning, demonstrating a high degree of responsiveness and technical competence. The project culminated successfully, with the resultant model being later utilized in research work, thereby validating the accuracy and applicability of the developed model.

The "Electrostatic Radon Behavior" project was an insightful investigation into the patterns of radon and its progeny accumulation on electrostatically charged surfaces, such as sweaters. One of the primary challenges was the difficulty of collecting samples in a reproducible manner due to the usage of balloons. This was addressed using a unique combination of charged balloons, time series gamma spectrometry, GM counting, alpha spectrometry, and alpha/beta discrimination. The study stood out as the first in-depth analysis of radon and radon progeny deposition on a latex balloon. It led to the pivotal discovery that the emanations largely originated from radon progeny rather than radon itself, significantly contributing to the understanding of radon behavior and its implications on electrostatic surfaces.