Winners

This project explores how a wearable tail could help seniors with Parkinson’s disease maintain balance, reducing their risk of falls. Using built-in sensors to detect when they are unstable, the tail acts as a counterweight, actively shifting their balance. Unlike many common support devices, the tail shines for its compact, modular design that goes beyond traditional structural belts, which cannot respond in real time. Testing yielded data demonstrating the tail's practical effectiveness in real-world settings, improving recovery from imbalances. Optimal configurations vary among individuals, enabled to be tested by the high customizability of the tail's 3D-printed components. The practical application of the tail holds the potential to make a great difference for people with Parkinson's: a low-cost, wearable solution reduces fall risk and keeps money in an individual's own pocket, allowing them to walk naturally feeling more independent and safe in daily life. Our Projectboard.

Less

When my father was hospitalized, overworked staff struggled to check his vitals regularly. That experience led me to remote heart-rate monitoring, then to a deeper problem: optical heart-rate systems falsely give normal readings on darker skin. Black patients experience three times the rate of undetected hypoxemia, contributing to 41% increased mortality. This bias was first documented in 1990, and despite decades of research, it remains unresolved. Most approaches focused on improving signal-to-noise ratio, but I returned to the underlying physics. I discovered a second pulsatile component at cardiac frequency that had been overlooked, and proved it is a dominant cause of demographic error. This led to three innovations: a method that improves equity by removing this signal component, a framework that separates coupled vascular-signals, and a three-wavelength oximetry system reducing pulse-oximetry bias from 2.3% to under 0.15%, without training data. The problem was never unsolvable; it was unexamined. My Projectboard.

Less

Heart disease currently affects roughly 6 million Canadians, with emerging research indicating that up to one-third of individuals with underlying cardiometabolic conditions will develop myocardial fibrosis. Fibroblast-targeted CAR-T cell therapy has emerged in preclinical studies as a highly promising method to actively regress scar tissue and restore heart function. However, clinicians currently lack a non-invasive tool to spatially monitor heterogeneous fibroblast clearance. My project addresses this research gap through a proposed non-invasive, circuit-based array to acquire multichannel cardiac acoustics. A beamforming algorithm is proposed, reconstructing irregular ventricular wall vibrations caused by fibroblasts into localized acoustic maps based on chaotic scattering. By leveraging an in vitro hydrogel phantom with graded chemical crosslinking to simulate fibrotic stiffness, the proposed non-invasive device validates its ability to detect localized acoustic signatures and quantify their impact. My Projectboard.

Less

Anorexia Nervosa is a severe psychiatric disorder that primarily impacts teenage girls. This condition is especially challenging due to the lack of concrete biomarkers, such as structural changes to specific brain regions or altered connectivity pathways. Furthermore, this project investigates the neurobiological effects of Anorexia Nervosa to determine a set of affected areas that can improve our diagnosis and treatment of future patients. Following the standard meta-analysis procedure, numerous studies were taken from across scientific databases, then screened according to strict inclusion criteria. fMRI research on resting-state connectivity and MRI research of gray and white matter volumes were the target experiments. The final set of studies underwent a program-based analysis to isolate clusters within the brains of patients that show abnormal volume or activation levels compared to healthy controls. In this way, the project synthesizes findings from across neuroimaging to create a representative model of the anorexic brain. My Projectboard.

Less

Nearly 50% of IVF embryos arrest before they become viable pregnancies, and the molecular mechanisms causing this failure remain poorly understood. The Digital Embryo is the first multi-omic framework integrating 6 omics — transcriptomics, proteomics, metabolomics, secretomics, genomics, and epigenomics — into a shared molecular-state system spanning 1,963 cells. A machine learning model trained on these omics predicts embryo arrest, achieving an AUC of 0.870 on unseen datasets. Beyond prediction, the Digital Embryo can map any individual embryo as a one-to-one molecular replica in silico. Its perturbation engine then allows for the modification of that embryo’s omic profile, culture conditions, and environmental exposures, and recalculates its arrest risk. Validated against 85 compounds, 37/40 of the strongest predicted effects on arrest matched known in vitro results. The Digital Embryo also generated 35 candidate treatment combinations against arrest, demonstrating the capacity for computational treatment discovery and personalized, multi-omic-guided reproductive medicine. My Projectboard.

Less

Sustainable Carboxylated Cellulose Hydrogel Beads for Pb(II) Wastewater Treatment Lead pollution is a major environmental and health problem, and current industrial treatments are damaging to the environment in both production and end-of-life disposal. In this project, CeLlulosE oxALate hydRogel (CLEAR) beads were developed as a novel, biodegradable, and highly effective solution for removing lead. CLEAR beads are made from natural substances found in plant fibres and algae, which lets it act like a “chemical magnet” that pulls lead out of water and removes it. The material can remove large amounts of lead in a matter of minutes and can be reused many times. By optimizing how it is made, CLEAR beads are designed to remain durable during use, decompose after use, and be low-cost to produce, while having strong potential to use wheat straw as its main raw material. Overall, CLEAR beads are a promising technology that opens new possibilities for greener heavy metal wastewater treatment. My Projectboard.

Less

1 out of 4 patients die within 3 months of a bloodstream infection from Staphylococcus aureus (S.aureus) [1]. Currently, both first-line and even last-resort antibiotics are mainly ineffective against infections by S.aureus because of its resistant mechanisms and stress response pathways that enhance bloodstream survival. To address this, I developed a bioactive hybrid treatment which targets multiple stress pathways in contrast to just one, compared with current treatment options. My treatment combines an Antimicrobial Peptide(AMP): DJK-5, with specific compounds derived from discarded Humulus lupulus(Hop) plants. Aiming to not only disrupt and eradicate the infection, but also to minimize the unwanted side effects current drugs inflict on patients, in order to break the toxic cycle of bacterial resistance. While this research is a work-in-progress towards clinical applications, this combination treatment has the potential to improve the long-term outcomes and quality of life for patients battling resistant infections. My Projectboard.

Less

Deafblindness is a combined disability of vision and hearing impairments that create significant challenges to communication and information access. Many deafblind individuals rely on intervenors to access spoken information, but access to these services is highly limited due to shortages, uneven availability, and funding restrictions. To address this gap, NEXUS was developed as a wearable assistive device that enables real-time, independent communication. NEXUS captures surrounding speech, identifies who is speaking, and converts conversations into accessible formats like braille and enlarged text, allowing users to follow group conversations without relying on human intermediaries. Notably, the device uses speaker diarization that combines spatial awareness with voice pattern analysis, making speaker identification robust even in complex, multi-speaker environments. This approach outperforms current state-of-the-art models for real-time, far-field diarization. Overall, NEXUS provides an accessible and scalable solution that enhances communication and supports greater autonomy among deafblind communities. Our Projectboard.

Less

By 2052, one quarter of the population will be over the age of 65, with adults over 85 the fastest growing segment. Supporting older adults to age well is a national priority. This project extends the CareBotix system into a platform-agnostic robotic architecture integrating manipulation, navigation, social interaction, multimodal AI. Manipulation was tested on object-handling tasks using the Isaac GR00T framework, achieving 93–100% task success and low joint-trajectory variance. Navigation used Isaac ROS for indoor localization and the Cosmos-Reason2-2B reasoning LLM for goal interpretation and adaptive planning; Isaac ROS built a stable map in 10 seconds, maintained 5–10 cm localization error, tracked 800–1,200 features per frame. Human–robot interaction used a multimodal MiniLM AI model achieving state of the art results in conversational AI latency and emotion classification. Results show a modular architecture is scalable while delivering robust manipulation, reliable navigation, naturalistic interactions-key for user acceptance. My Projectboard.

Less

The prevalence of microplastic pollution in aquatic environments is a significant global concern. However, quantifying microplastics using conventional laboratory-based methods is costly and labour-intensive, considerably limiting routine analysis. To overcome these constraints, I developed a field-portable, submersible, laser-based holographic imaging system for the in situ detection of microplastics. Digital holography was employed to facilitate high-detail, large depth-of-field imagery through an extensive water volume, achieving an optical resolution of <10 μm through optimization. Numerical algorithms were implemented to accurately reconstruct images and extract key morphological features of microparticles. Custom-trained AI models were developed to identify irregular-shaped microplastics in lake water, resulting in 94% detection accuracy. Performance was evaluated via real-world field trials at 10 bodies of water, yielding statistics on particle characteristics and concentrations. In addition to microplastics, various microorganisms were identified in situ, demonstrating the system’s versatility as an innovative, low-cost platform for robust, real-time monitoring of ecosystem health. My Projectboard.

Less

Oral diseases affect nearly 3.7 billion people worldwide, making them one of the planet’s most widespread and neglected health burdens. Although largely preventable, these conditions often progress unnoticed due to late detection, and accessibility. In my project, I am researching, developing, and pilot-testing a low-cost early-detection breath analysis device that screens for biomarker patterns most associated with cavities, gingivitis, halitosis, diabetes, and other oral or systemic health risks. I constructed BART, a low-cost, battery-powered wireless breath analyzer built on an ESP32-S3 microcontroller and five metal-oxide-semiconductor (MOS) gas sensors, with all analysis performed on a Raspberry Pi 5 hub. Through gain-offset calibration and environmental consideration, and time-series feature extraction and classification algorithms B.A.R.T generates a risk score supported by front-end display features and back-end data analysis/logging. B.A.R.T allows early detection, accessibly, and non-invasively, advancing breath biopsy. Frequent home-based self-screening can enable early identification of dental disease risks, thus facilitating timely interventions. My Projectboard.

Less

The plastic mulch film, a plastic sheet applied to agricultural soils, increases crop yield through its benefits such as moisture retention and weed suppression. However, due to their potential of forming microplastics, many biodegradable alternatives have emerged. However, these alternatives still showcase a high leaching potential of toxic substances or substances with unknown environmental fate (Michaela K. Reay, 2025). I aimed to address this gap by asking a question. Can a novel polysaccharide-based mulch film serve as a lower leachate toxicity alternative to conventional films while preserving performance? The film showcased lower leachate toxicity under the testing conditions due to likely lack of inhibition in bioassays, hydrophilic functional group presence, and low material toxicity. The film also met performance by meeting/exceeding many physical and mechanical properties. This project demonstrates the potential for increased crop yield without compromising soil, plant, or even human health. Ultimately, providing a more sustainable option. My Projectboard.

Less

Many patients are injured when IV fluids accidentally leak into their skin instead of staying in the vein. This causes painful swelling and bruising, and it happens to nearly 1 in 10 people. To solve this, I developed Veino, a wearable device equipped with smart pressure sensors. Unlike current methods where nurses have to check for leaks by sight, Veino monitors the IV site automatically. It detects tiny pressure changes and sends an early warning before a serious injury can occur. I also created Veino Pro, which uses these same "hero" sensors to monitor pressure inside the body during medical procedures like endoscopies. By catching these issues early, my technology helps healthcare workers act faster and prevents unnecessary pain. This project makes hospital treatments safer for millions of people by turning a common medical problem into a smart, digital solution. My Projectboard.

Less

Many people struggle to take medications correctly, leading to missed doses and potential hospitalizations. Existing pill dispensers require someone to read each label, hand-load pills into compartments, and count out doses, making them hard to use for people with disabilities who don't have caretaker support. To solve this, I designed Pharma-Bot, an AI-enabled self-dispensing pillbox with two components: a software program and a 3D-printed pill dispenser. The software, written in Python, uses a camera and Optical Character Recognition (OCR) to extract key information from prescription labels and tells the dispenser which pills to release, how many, and when. The dispenser has three compartments and uses an Arduino-controlled servo motor to release the correct number of pills. In testing, Pharma-Bot read labels and dispensed pills with high accuracy. By combining smart label reading with automatic dispensing, Pharma-Bot is a low-cost solution that helps people safely manage their prescriptions at home. My Projectboard.

Less

2.3 million women are diagnosed with breast cancer every year, world wide. One of the main tools to decide the treatment is tumor grade, or aggressiveness of the cancer cells. Yet, the pathologists agree only about 50-60 % of the time on grade scores for these cancers. I have created an AI powered tool, C-Strik to assist pathologists in grading breast cancer consistently using objective criteria, for all the women, regardless of where they are. Additionally, my ultimate aim is to optimize C-Strik to be used as a robust alternative for multigene tests in breast cancer. These tests are used to decide whether a woman will benefit from chemotherapy or not, however, they each cost 5000 CAD and are not accessible in many parts of the world. C-Strik can be an inexpensive and reliable alternative so that no woman is deprived of their optimal treatment choice, ever! My Projectboard.

Less

Over 98% of potential therapeutics for neurological diseases fail because they cannot cross the blood-brain barrier (BBB). Current screening methods for novel ligands often focus too heavily on binding affinity, which can lead to the "Transferrin Paradox," where stronger binding reduces downstream release into the brain. To address this, I developed EXODEC, a computational framework that evaluates ligands across five sequential stages: adherence, binding, uptake, transcytosis, and release. EXODEC now goes beyond static scoring by integrating a BBB transcytosis atlas, a context-aware digital twin, safety screening, payload-aware modeling, cross-species design, autonomous ligand discovery, dynamic pharmacokinetic simulation, and a geometry-aware molecular docking layer. Using this platform, I identified why benchmark ligands fail, mapped residue-level design landscapes, and refined EXODEC-family peptides such as EXODEC_GEN_001 for strong predicted delivery performance. EXODEC provides a scalable platform to de-risk brain-targeted drug development and accelerate therapeutic design for diseases including glioblastoma. My Projectboard.

Less