This capstone investigates a Public Safety Broadband Network (PSBN) prototype that integrates Band 14 LTE with complementary networks and evaluates the use of multipath protocols, including MPTCP, to improve throughput, reliability, and adaptability. The project focuses on whether TCP-based applications remain usable under disruptions such as link loss and handovers.

This project presents a data-driven framework for analyzing atmospheric electrification using airborne microphysics, radar observations, and lightning activity. The system integrates multiple data sources into a unified Electrification Risk Index (ERI) that quantifies storm intensity and charge separation potential. Using high-resolution particle imagery from optical array probes (2D-S and HVPS-3), ice particle orientation and microphysical properties are extracted and combined with lightning flash rates and atmospheric dynamics to characterize electrified regions within convective storms. The resulting ERI enables both spatial and temporal analysis of storm electrification, highlighting high-risk regions and capturing the evolution of electrical activity throughout the storm lifecycle. This platform serves as both a visualization tool and an analytical framework, demonstrating how microphysical observations can be directly linked to lightning behavior and used to estimate electrification risk in real time.

In collaboration with TELUS Communications In this project, different fibre deployment designs for low-density and rural locations are investigated. In sparsely populated areas, traditional fiber-to-the-home (FTTH) deployments are expensive because of lengthy fibre runs and redundant infrastructure.

Free-Space Optics (FSO) is a technology that uses optical/laser light propagating in free space to transmit data—think of it as "fiber optics without the fiber." Primary Application: Ideal for high-bandwidth, point-to-point links ("Last Mile") where laying physical cables is costly. The system is designed to enhance precision and reliability in optical communications involving UAVs, drones, and fixed targets. A closed-loop control mechanism utilizing a Spiral Scan Algorithm continuously monitors and adjusts the MEMS mirror orientation in real time, achieving sub-5 second lock-on speeds. Our machine vision tracking system, powered by the MKR 1310 with sensor polling every 2 seconds, maintains 85% signal intensity during indoor testing—demonstrating robust performance for drone and fixed target communication.

This project compares RGB-only, thermal-only, OR late fusion, Product-of-Experts, and intermediate feature fusion models for wildfire detection. The goal is to make wildfire detection more reliable when smoke, low visibility, or hot clutter make a single sensor unreliable on its own.

Autonomous vehicles rely heavily on accurate perception systems to understand their surroundings and make safe decisions. One of the most critical components of these systems is object detection, which enables a vehicle to identify and localize surrounding objects such as vehicles, pedestrians, and infrastructure. However, reliable object detection remains challenging in adverse conditions, where lighting, weather, and scene complexity can significantly degrade sensor performance.

Structured light provides new ways to control and probe optical systems. In this capstone project, we investigate how polarization vortices interact with structured matter, ranging from twisted optical paths to optically active molecular media. Through theory, simulations, and experiments, we demonstrate that structured polarization can significantly enhance observable rotation effects, including a doubling of rotation angle compared to conventional linear polarization.

Laser beam shaping is essential in modern day photonics, but traditional optics setups are static, expensive and challenging to modify. Our project introduces a novel "Three-for-One" architecture capable of real-time swcithing between three distinct and popular beam modes: Gaussian, Flat-top, Optical Vortex. By leveraging orthogonal polarization states implemented using a polarization controlled Mach-Zender Interferometer (MZI) , we have designed a compact, cost effective system that allows for reconfigurable beam profiles on demand.