Celebrating Our Graduates

The School of Graduate and Postdoctoral Studies (SGPS) is proud of the accomplishments of our graduates and excited to showcase the diverse and impactful research they have undertaken and acknowledge their contributions to Ontario Tech's rich culture of innovation and excellence. To celebrate their achievements, we are pleased to present a selection of our recent graduates and details regarding their completed studies.

Ainaz Alavi
Master of Science in Computer Science
My research focused on whether we can trust AI in a field as sensitive and important as bloodstain pattern analysis. I explored different AI methods being used in the field and also spoke with experts to understand their perceptions, experiences, and opinions about using AI in bloodstain pattern analysis. Through this program, I learned how to conduct interdisciplinary research and gained a better understanding of the challenges, importance, and necessity of bringing together different fields and perspectives when working on complex research questions.

Gillian Slade
Master of Health Sciences (MHSc) in Kinesiology
Overhead work is common in many occupations and can contribute to shoulder fatigue and musculoskeletal injury. Gillian’s thesis investigated how the pace of repetitive overhead work influences muscle fatigue, endurance time, and perceived workload. By comparing four work cycles that varied how frequently workers performed repeated overhead movements, while maintaining the same overall work-to-rest ratio, her research found that faster cycles substantially reduced endurance times and increased perceived workload. These findings demonstrate that work-to-rest timing can influence fatigue, even when overall proportions remain the same, and can help inform the design of fatiguing occupational tasks.

Guadalupe Blanco Velasco
PhD in Forensic Psychology
False confessions can be as detailed as presumed true ones, but whether they differ in broader language patterns is unclear. This dissertation compared 20 false and 20 presumed true confessions in U.S. police interview transcripts, using large language models to analyze communicative intent and topic by speech unit. Police officers and false confessions followed a more constrained style, with more closed-ended questions. Suspects' contributions in false confessions were shorter and more dependent on prior speech units, while presumed true confessions showed more affective and expressive content. Though not yet diagnostic, findings may inform research on false confession profiles.

Hind Isleem
Master of Software Engineering
My research focuses on intelligent control and data-driven modeling for IoT-enabled systems. My thesis develops PoolPilot, a gray-box Model Predictive Control (MPC) framework for autonomous water chemistry management in commercial swimming pools. Using real-world IoT sensor and dosing data, the research combines system identification, augmented Kalman filtering, and predictive control to regulate pH and oxidation-reduction potential (ORP). The work explores how existing IoT infrastructure and historical operational data can move beyond monitoring toward safer, proactive, and autonomous control.

Jen Calver 
Doctor of Philosophy in Health Sciences
The overarching aim of this thesis was to explore psychological health and safety for workforce retention in long-term care (LTC) homes. This study explored psychological health and safety workplace factors for workforce retention (PHS WR) in LTC. Findings identified 45 PHS WR priorities that achieved consensus from a panel of LTC leaders, and identified 24 implementation characteristics with 59 barriers and 61 facilitators. Findings demonstrate the complex nature of staffing stability and succession planning in LTC, and proposes strategies and information to guide organizational leaders in decision-making and operational management within the system, the workplace, and interactions with individuals.

Krystina Clarke
Doctor of Health Sciences
My doctoral research contributed to the development of STEPS (Simulation Technology, Educational Principles, and Safety), Canada's first undergraduate Minor designed to prepare Simulation Operations Specialists, the non-clinical professionals who run healthcare simulation. Across three studies, I mapped the occupational foundations of this emerging workforce, assessed the feasibility of launching the program, and designed a competency-based curriculum blueprint. This work helps formalize simulation operations training for a recognized health professions career while strengthening the teams behind safer patient care. I'm so grateful that my 10 years at Ontario Tech wrapped up with such meaningful, impactful work!

Kushan Lulbadda Waduge
Doctor of Philosophy (PhD) in Electrical and Computer Engineering
Kushan’s research presents a solid-state transformer (SST)-based EV fast-charging station integrated with a comprehensive multi-layer protection framework. The charging system improves fault tolerance, enables flexible charging, and reduces downstream converters. The coordinated protection framework combines ultrafast short-circuit interruption, overcurrent and overvoltage protection, and harmonic-analysis-based detection of internal open-circuit semiconductor switch faults. Rogowski-coil and Hall-effect sensors are integrated with the system to minimize implementation cost. The smart solid-state circuit breaker implemented in the research achieved a fault response time of 180 ns, demonstrating ultrafast performance. The findings demonstrate a low-cost, high-speed, and intelligent protection framework for next-generation electrified power networks.

Matthew Oinonen
PhD in Electrical and Computer Engineering
I integrated Explainable Artificial Intelligence (XAI) with cyberattack detection in the Smart Grid to develop more trustworthy detection schemes. I performed a novel sensitivity analysis to analyze several XAI methods and developed guidance on what XAI methods to use and how to select their parameters. Furthermore, I developed a scheme to utilize custom wavelets to improve the detection accuracy of a Smart Grid cyberattack detector. I tested my proposed methodologies on datasets generated from a Hardware-in-the-Loop setup in my research lab that implements Smart Grid systems.

Md Hasan Saju
MASc in Software Engineering
My thesis presents an end-to-end framework powered by large language models (LLMs) that brings threat classification, risk-based prioritization, automated evidence collection, and incident resolution support together in one pipeline. The framework combines machine learning, multiple LLMs, and ensemble decision-making to identify critical threats, generates executable investigation queries for leading security platforms, and recommends resolution steps. Evaluations show improvements over existing approaches, helping reduce analyst workload and enabling faster, more consistent security decisions. This work moves security operations toward greater autonomy, helping organizations respond to cyberattacks more quickly and ease the growing shortage of skilled security analysts.

Melissa MacMillan
Doctor of Philosophy, Community, Public and Population Health, Faculty of Health Sciences
My thesis evaluates the impact of using common methodological trade-offs in rapid reviews compared to the accepted standard in systematic reviews. My results add to the understanding of whether rapid review results can be trusted. They also provide practical knowledge on how rapid review authors can reduce the potential biases when selecting and using methodological trade-offs to conduct rapid reviews, ensuring the evidence provided to decision-makers align with interest-holders' needs and provide trustworthy evidence to inform policies.

Mohammadreza Saberironaghi
MASc, Electrical and Computer Engineering
My research focuses on developing an efficient deep learning model for automated steel surface defect segmentation. The research introduces ID-MSNet, an enhanced version of the UNet3+ architecture, designed to accurately identify and segment different types of steel surface defects, including inclusions, patches, and scratches. The proposed model combines multi-scale feature learning, improved down-sampling, and convolutional attention mechanisms to better capture defects of different sizes and characteristics. The goal of this research is to improve the accuracy and efficiency of automated visual inspection.

Nolan Ford
Master of Health Science
This research examined how ergonomic gaming mouse design affects the physical demands and performance of competitive video game players. Collegiate esports players completed aiming and gameplay tasks using both a conventional gaming mouse and one designed with additional palm support. Wrist and arm movement, muscle activity, perceived discomfort, usability, and gaming performance were assessed. The findings suggest that palm-supported mouse designs can improve wrist posture without compromising overall performance, but reducing physical strain requires considering gaming task demands rather than relying on mouse design.

Sakil Sarker
MSc in Computer Science
My thesis presents CoRead, a mixed-initiative collaborative reading interface that tackles two open problems: when AI should intervene, based on sustained behavioural signals, and how help should be routed across individuals, peers, and the group. A formative study (N = 12) produced five design requirements, which shaped a coordination architecture spanning pre-reading alignment, in-session support, and post-session synthesis. It abstains when intervention would disrupt more than help. A Wizard-of-Oz evaluation (N = 12) showed 81.1% acceptance of implicit AI interventions, 53.8% acceptance of peer-routing offers, SUS of 73.3, and NASA-TLX of 47.0.

Sandra Koshy
PhD in Mechanical Engineering
My thesis presents a binder-free approach for fabricating transition metal oxide electrocatalysts using Ultra-Short Laser Pulses for In-Situ Nanostructure Generation (ULPING) for enhanced water splitting. Laser processing was used to directly engineer nanostructured electrode surfaces, with studies examining different transition metal oxide materials and laser processing conditions. The resulting electrodes were evaluated for their electrochemical performance in water splitting, focusing on catalytic activity, charge transfer, surface characteristics, and stability. The research demonstrates the potential of ULPING as a scalable approach for developing durable and efficient electrocatalysts for green hydrogen production.

Sonia Mirza
Master of Health Sciences - Community, Population, and Public Health
To date, dementia does not have a cure, relying on preventive strategies to mitigate its significant impact on patients, communities, and healthcare personnel. With this call to action, Sonia's study explored anemia as a modifiable risk factor. She conducted a cross-sectional study using patient charts and found a considerable proportion of anemia among patients with dementia. Further, the study describes potential high-risk groups by sex, age, and comorbidities. As the first Canadian study on this relationship, Sonia aimed to set the foundation for larger-scale studies to tackle this public health crisis. 

Theresa Watt
Doctor of Education
As both a teacher educator and former classroom teacher, I have always been interested in assessment that feels meaningful. My research focused on authentic assessment in teacher education and how assessment tasks can better prepare pre-service teachers for the complex, practical, and reflective work of teaching.

Tina Cloutier
Master of Health Sciences - Health Informatics
My research focused on asynchronous data collection for remote health and wellness interventions. Working with Dr. Judith Andersen's Autonomic Modulation Training (AMT) study, a home-based psychophysiological resilience training for police officers, I extended Dr. Carolyn McGregor's Big Data analytics framework, instantiated in the Artemis and Athena platforms, to decouple data collection from data acquisition. This allows physiological data to be gathered from officers in their own homes rather than in synchronous, in-person lab sessions. The work addresses scalability limits in resilience training research and supports efforts to reduce the risk of posttraumatic stress injuries among police.

Troy Radam
Master of Applied Science - Mechanical Engineering
My research focuses on autonomous heterogeneous multi-robot systems, specifically cooperative operation between Unmanned Aerial Vehicle (UAV) and Unmanned Ground Vehicle (UGV) robots for use in complex environments. The work in my thesis presented a design and pipeline that would enable the autonomous deployment and stowing of a quadcopter from a Boston Dynamics quadruped robot. The platform was designed with passive geometry to minimize the use of active components, enabling repetitive deployments in a GPS-denied environment. The system is capable of transporting the quadcopter in terrain that is accessible to the quadruped robot utilizing an Electro-Permenant Magnet (EPM).

Wahab Hassan Janjua
Master of Applied Science in Mechanical Engineering (Mechatronics)
My research focused on improving the accuracy of handheld 3D position tracking using an Inertial Measurement Unit (IMU). IMUs are small sensors that measure motion, but their position estimates can quickly develop errors. I developed a bio-inspired method to detect short moments when the sensor is stationary and use these moments to reduce tracking errors. I also used machine learning and a small camera to further improve position accuracy. The goal was to make handheld motion tracking more accurate, practical, and reliable.