Blue Sky Solar Racing

I was the Chief Structural and Manufacturing Engineer for the University of Toronto’s Blue Sky Solar Racing Team, where we designed and raced solar-powered vehicles across a grueling 3,000+ km course through the Australian Outback in the World Solar Challenge. Our mission centered on creating an efficient solar car, while promoting renewable energy and sharing our expertise with fellow students.

As the Chief Structural and Manufacturing Engineer, I led the fabrication team through the design and construction of the car’s core components, with a focus on optimizing performance and efficiency. I was responsible for streamlining manufacturing processes, such as reducing seamlines during the plug-making phase, which resulted in a 17% reduction in material waste and improved overall production speed.

I also oversaw the infusion of carbon fiber and kevlar layups, performing tests to refine the flow rate and ensure the car’s aerodynamic flexibility. My role extended to managing the team’s 3D printing operations, where I optimized print profiles to reduce time by 20% while maintaining quality.

Throughout the project, I worked closely with other sub-teams—electrical, strategy, and solar array—to ensure structural and manufacturing goals aligned with the overall design and performance objectives. These experiences have not only honed my technical expertise but also strengthened my leadership and problem-solving skills, culminating in our successful participation in the World Solar Challenge.

Robotics for Space Exploration

I was a core software and autonomy member of the Robotics for Space Exploration Design Team, participating in competitions such as the Canadian International Rover Challenge (CIRC) and the University Rover Challenge (URC).

During my time on the team, I gained hands-on experience with sensor implementation and control systems using ROS and Python. For CIRC 2022, I was responsible for setting up and programming the lidar and IMU systems on the rover. This involved understanding technical specifications and collaborating with the electrical team to optimize power usage without compromising visibility. The lidar proved vital during the science task, where it provided a detailed topographical map that improved reporting accuracy by 20%. I also ensured the functionality of the rover’s sensors and motors and contributed to the Arduino code for the rover’s arm.

Additionally, I worked on setting up a Ubuntu server on our old base laptop and researched autonomy solutions to advance the rover’s performance in future competitions

Formula 1 Predictor

As a lifelong Formula 1 fan, having watched the sport since I was 6, I am currently working on a personal project to develop a machine learning algorithm that predicts the outcome of upcoming races. This project blends my love for F1 with my passion for data analysis and machine learning, using a variety of tools and techniques such as Python, pandas, and TensorFlow to analyze large datasets that include key factors like driver performance, team statistics, and track history.

To build a robust predictive model, I’m using Keras Tuner for hyperparameter optimization, which allows me to fine-tune essential parameters such as learning rates, dense layer units, and dropout rates. This ensures the model is optimized for performance and accuracy. Additionally, I'm applying regularization techniques like BatchNormalization and Dropout to prevent overfitting, making the model more generalizable.

I’m also incorporating recency bias using a decay function, which weighs recent race performances more heavily to ensure the model remains relevant to the current season’s trends. This, combined with feature engineering, allows me to refine critical metrics such as average driver position, team performance, and track-specific history, giving the model deeper insights into performance patterns.

To ensure the model trains efficiently, I’m using callbacks like EarlyStopping and ReduceLROnPlateau, which dynamically adjust the training process based on validation loss and prevent overtraining. Additionally, I implemented a custom loss threshold callback that halts training once a desired validation loss is achieved, improving overall efficiency.

Ultimately, my goal is to develop a model that not only predicts race outcomes with a high degree of accuracy but also provides valuable insights into performance trends within Formula 1. This project challenges me with tasks like data normalization, hyperparameter tuning, and model evaluation, while continuously fueling my passion for both machine learning and the sport I’ve loved since childhood.

Engineering Without Borders

During my time with Engineers Without Borders (EWB), I led the Curriculum Change Project, which aimed to incorporate more sustainability into the Engineering Strategies & Practice (ESP) course at the University of Toronto. Initially, the project’s scope was broad, but after taking over as lead, I rescaled the focus to concentrate on making impactful changes within the ESP course.

I recruited new team members, conducted in-depth research by reviewing similar initiatives at other universities, and partnered with the Sustainability Engineering Association to launch a university-wide survey. This survey helped us gather insights from engineering students on how sustainability could be better integrated into the curriculum.

The culmination of our efforts was a comprehensive report, detailing actionable strategies for incorporating sustainability into the course. I also engaged with faculty members to discuss the best approaches for implementing these changes. Though I’ve since stepped down from my leadership role, I continue to contribute as a consultant.

This project significantly enhanced my communication, leadership, and research skills, as well as my ability to work with diverse teams toward a common goal.

Page updated

Google Sites

Report abuse