Innovative Projects Realized

In-vitro digestibility, bioactive peptide generation and functional characterization of lupin protein isolates modified by high-voltage atmospheric cold plasma

This project explores how different processing methods affect the nutritional quality and health benefits of proteins from lupin, a sustainable legume crop. Lupin is rich in protein, environmentally resilient, and cost-effective to produce, making it an attractive alternative to meat and dairy proteins. However, the way proteins are extracted and processed can strongly influence how well they are digested and what health-promoting compounds they release.
The project will compare two extraction methods: the conventional alkaline method and an innovative green approach using deep eutectic solvents (DES). Some protein isolates will also be treated with high-voltage cold plasma (HVCP), a novel non-thermal technology that can enhance protein functionality. These proteins will then be tested in a dynamic in vitro digestion system that mimics human stomach and intestinal conditions. The study will identify and measure peptides released during digestion and evaluate their potential health benefits, including antioxidant and blood pressure-lowering activities.
By linking extraction and processing methods to digestibility and bioactivity, this research will support the development of high-value plant-based food ingredients. The collaboration between the University of Manitoba and the University of Hannover will also strengthen international partnerships, foster trainee exchanges, and create opportunities for future joint projects.

Deep Learning-Based Classification of Defects in Induction Brazed Copper Joints Using Infrared Thermal Imaging for Enhanced Joint Quality Monitoring

This project will develop a new system that uses infrared cameras and artificial intelligence to automatically detect defects in copper joints made by induction brazing, a process widely used in heat pump manufacturing. By capturing and analyzing the heat patterns that appear during brazing, the system can quickly identify problems such as cracks or poor filler flow without the need for expensive and time-consuming testing. The collaboration between the University of Manitoba and the University of Aveiro will combine expertise in smart manufacturing and brazing technologies to create a prototype real-time monitoring tool. This will help industry partners improve the quality and safety of heat pumps, reduce costly rework, and support the transition to more energy-efficient and sustainable heating and cooling systems.

Optimization and Scale-Up of a Biomanufacturing Process for Anellovirus-Based Vectors for Gene Therapy Applications

Gene therapy is a cutting-edge treatment that can fix or replace faulty genes to treat diseases, but producing the necessary gene delivery cargo, called vectors, is currently expensive and complex. Many of today’s vectors also trigger immune responses in patients, limiting their efficacy. This project explores a new type of vector made from a harmless human virus called anellovirus, which exhibit natural immune evasion properties, indicating a potential for repeat dosing and long-term therapeutic applications without causing immune responses. The goal is to produce these anellovirus-based vectors using bacteria instead of animal or human cells. The partner company, Theraphage Inc., holds the rights to a unique technology called the iPhAGE system, which engineers bacteria to produce anellovirus-based vectors. Partnering with Theraphage, the interns will help optimize and scale up the biomanufacturing process for producing these vectors. This will reduce manufacturing costs, support the development of safer gene therapies, and create new job opportunities in Canada’s growing biotechnology sector.

2026 FIFA World Cup Marketing Plan Evaluation for Ontario Soccer

This project will help Ontario Soccer make the most of the upcoming 2026 FIFA World Cup by evaluating how well their “GeneratiON26” marketing plan is working and suggesting ways to improve future efforts. With games being hosted in Toronto, the World Cup is a unique chance to grow interest in soccer across Ontario. A Mitacs Business Strategy intern will work with Ontario Soccer to assess the success of its FIFA-related marketing activities, and recommend new ideas to help more people get involved in the sport. The project will give Ontario Soccer useful tools and insights to guide future marketing and community outreach efforts beyond the World Cup.

Optimizing Steam Pyrolysis Parameters for Biochar Production from Invasive European Buckthorn

This MITACS research project tackles two critical environmental challenges: the costly removal of invasive plants and the improvement of degraded urban soils. Municipalities currently invest substantial resources in removing invasive European buckthorn, which is often burned or sent to landfills, releasing stored carbon back into the atmosphere. Additionally, many urban soils are compacted and lacking in organic carbon, which impairs both water infiltration and soil respiration.

The project will expand opportunities for Senti Solutions’ Steam Pyrolysis technology, using it to convert European buckthorn into biochar—a charcoal-like substance that improves soil health. Unlike traditional burning methods, steam pyrolysis offers precise control over temperature and processing time, allowing for the creation of biochar with tailored properties. This targeted biochar can be particularly beneficial for urban soil restoration, as it improves soil water retention and aeration, both of which are essential for the growth and maintenance of trees in cities.

The MITACS intern will spend four months at Senti’s Guelph facility testing various processing conditions, measuring energy efficiency, and analyzing the quality of the resulting biochar. This research will establish production protocols for biochar that meet international standards for soil amendments, while also rigorously documenting the commercial viability of this waste-to-value approach.

Towards Real-time Assessment of Drift Conditions during Potash Mining

The safety of underground mining operations depends heavily on monitoring and management of various hazards that can arise within mines and their supporting infrastructures, particularly the drifts (tunnels). Drifts are critical pathways within mines, and their structural integrity is paramount to ensure the safety of workers and equipment. Traditional methods of evaluating drift stability often involve manual inspection and/or manual collection of data from local instrumentation clusters. This can be time-consuming and limited in scope. The proposed research will explore the use of artificial intelligence (AI) technologies to assess the structural integrity of the drifts in the potash mines. It involves assessing the parameters and criteria required to identify the precursors to failure, as well as the means to collect those parameters in real-time. The findings from this research are expected to offer insights into the implementation of advanced AI solutions for hazard detection, maintenance scheduling and risk mitigation in mines.

Analysis of the electricity demand profile of a ski resort: characterization and identification of demand management opportunities

In recent years, Québec’s ski resorts have struggled with mounting electricity expenses. Warmer winters and growing stress on the provincial grid have driven up the energy needed for snow-making, particularly in the run-up to the holiday season. This significant cost is primarily driven by demand charges incurred from the intensive use of snow-making equipment prior to operating hours. As a result, ski resorts often spend tens of thousands of dollars on snow-making, which can account for 17% to 38% of their total electricity bill. This issue is expected to intensify in the near future, placing additional financial and operational strain on ski resorts in Québec and across Canada. While some ski station operators have begun exploring alternative control strategies and new technologies, there remains a pressing need for dedicated research into load management solutions for this sector. Previous studies have assessed the overall impact of snow-making at various resorts and documented mitigation strategies. The proposed project aims to: (a) conduct an on-site assessment of the electric load profiles of key circuits (snow-making, ski lifts, lighting) at a large ski resort, and (b) identify practical measures to reduce the resulting aggregated load.

Apprentissage machine quantique pour la prédiction de l’efficacité de nouveaux médicaments anti-inflammatoires : application pour la prédiction de l’affinité d’inhibiteurs de l’enzyme 5-lipoxygénase

Le développement d’une molécule de la conception jusqu’à l’obtention d’un médicament est un processus couteux. Dans ce processus, la première étape consiste à synthétiser une série de molécules qui inhibent une action biologique particulière, généralement médiée par un récepteur ou une enzyme. Les essais biologiques et cliniques peuvent être les étapes les plus couteuses de tout le processus de conception d’un médicament. Le recours à l’intelligence artificielle pour la modélisation moléculaire pourra réduire considérablement le temps et par conséquence les couts de cette importante étape de la découverte d’un médicament.
Parmi les cibles thérapeutiques d’intérêt, les lipoxygénases sont des enzymes cliniquement pertinentes qui sont responsables du métabolisme de l’acide arachidonique par l’insertion d’un atome d’oxygène. Parmi les pathologies liées aux métabolites générés par les lipoxygénases, on peut citer l’athérosclérose, l’asthme, la polyarthrite rhumatoïde et le cancer. En particulier, la 5-lipoxygénase (5-LO) semble être liée à ces maladies à composante inflammatoire, notamment les maladies des voies respiratoires telles que l’asthme et la rhinite.
Dans ce projet, nous proposons d’utiliser l’apprentissage automatique quantique pour simuler rapidement et avec précision l’affinité des inhibiteurs de la 5-LO. Cette approche représente une alternative prometteuse aux méthodes expérimentales traditionnelles, en accélérant la découverte de nouveaux traitements.

Hybrid AI Pipelines for Compression and Retrieval in Long-Term Financial Decision-Making

This proposal aims to address a growing challenge in Canada’s financial technology sector that is to analyze ultra-long-term, unstructured financial documents in private markets for explainable decision making. Canadian investors play a significant role in the global private capital market which is estimated by industries to exceed USD $20 trillion by 2030. Unlink public markets, where data is standardized and accessible, private investments rely heavily on thousands of long and inconsistently structured documents with redundant sections and content. Analysts spend 15–20 hours per week manually extract, verify and interpret KPIs and risk metrics from these documents. Top private-equity firms begin to adopt and deploy AI for speed, and cost-effectiveness but acknowledge the risks caused by current AI models that truncate long-term data, use simple aggregation, missing critical anomalies along long trends over decades. This opacity reduces transparency and trust, hence hampers investment decisions and risk economic inefficiencies in Canada’s private market which supports Canada’s pension funds. This project advances new technical knowledge at the interaction of AI, finance and software engineering. 1) Redundancy-Aware Token Compression. A method for compressing long and redundant financial documents while preserving multi-quarter, multi -year trends and identifying anomaly essential in decision making. 2) Multi-Resolution Document Analysis Pipelines. A processing pipeline that extracts key insights using key-value query framework developed byAltQ.ai at multiple timescales including sentence-level, section-level, document-level to support trend detection and anomaly alerts over long time span of decades. 3) Financial-Domain Metrics and Benchmarking. The definition of financial-specific metrics and indexing surpass SOTA’s limitation in uniformed chucking of long documents. setting the benchmark validates over thousands of datasets. 4) Explainable and Human-AI Collaboration Design. The design that embraces explanation and human-AI collaboration ensures compliance with Canadian and global regulatory frameworks for data and AI governance in finance.

Analyse sémiologique des stratégies d’oralité et des structures narratives dans Allah n’est pas obligé et Monkey Beach 

Ce projet de recherche porte sur l’analyse de deux romans, Allah n’est pas obligé de l’écrivain ivoirien Ahmadou Kourouma et Monkey Beach de l’autrice autochtone canadienne Eden Robinson. L’objectif est d’étudier comment ces auteurs utilisent des formes orales (comme les récits parlés, les proverbes, ou les dialogues familiers) et des structures narratives particulières pour raconter des histoires marquées par la guerre, la mémoire, la famille ou la culture. Grâce à une approche appelée « analyse sémiologique », le projet cherchera à mieux comprendre comment les mots, les sons et les formes de récit transmettent des émotions, des messages et des identités culturelles. Ce travail pourra servir à enrichir les outils d’analyse littéraire, à mieux valoriser les récits issus de contextes marginalisés, et à encourager une meilleure reconnaissance des voix culturelles diverses dans la société.

Approche d’Apprentissage Quantique pour la Découverte d’Inhibiteurs de la 5-Lipoxygénase : Vers de Nouveaux Agents Anti-inflammatoires

Le développement de nouveaux médicaments constitue un processus long, coûteux et complexe, s’étendant de l’identification de composés bioactifs jusqu’à leur mise sur le marché. L’une des premières étapes de ce processus consiste à concevoir et synthétiser des molécules capables de moduler une activité biologique spécifique, généralement en interagissant avec une enzyme ou un récepteur cible. Ces dernières années, l’intelligence artificielle (IA) a émergé comme un levier puissant pour optimiser et automatiser cette phase cruciale, en permettant une réduction significative des délais et des coûts associés à la découverte de candidats thérapeutiques.
Parmi les cibles enzymatiques d’intérêt, les lipoxygénases occupent une place importante en raison de leur rôle dans le métabolisme de l’acide arachidonique, catalysant l’incorporation d’un atome d’oxygène dans cette molécule. Les produits de cette réaction sont impliqués dans diverses pathologies inflammatoires et chroniques, telles que l’asthme, l’athérosclérose, la polyarthrite rhumatoïde ou certains cancers. En particulier, la 5-lipoxygénase (5-LO) joue un rôle central dans plusieurs processus inflammatoires, en faisant une cible thérapeutique de choix. Dans ce contexte, le recours à des outils informatiques pour générer de nouvelles molécules constitue une alternative précieuse aux méthodes expérimentales traditionnelles, en réduisant les coûts associés au développement de nouveaux agents thérapeutiques ciblant la 5-LO.

Next Generation Cybersecurity – Guiding Human-AI Collaborative Intelligence & Work Transformation in Security Operations Centre (SOC) – eSentire – U. Waterloo

Our research focuses on the critical intersection of artificial intelligence and cybersecurity operations, specifically examining how AI tools can be effectively integrated into Security Operations Centre (SOC) workflows. As cyber threats become increasingly complex, we investigate the delicate balance between leveraging AI automation and maintaining human expertise in threat detection and response.

Using Cognitive Task Analysis (CTA) as our primary methodological framework, we study how SOC operators process information and make decisions during security incidents. Our work addresses key challenges in human-AI collaboration, including automation bias, skill degradation, and the risk of reduced situational awareness that can occur when operators over-rely on automated systems.

We advocate for adaptive automation approaches where control dynamically shifts between human operators and AI agents based on task demands and context. This research is particularly relevant given recent advances in large language models for cybersecurity applications and growing concerns about operator burnout in high-stress SOC environments.

Our goal is to develop evidence-based guidelines for implementing AI support systems that enhance rather than replace human judgment, ensuring that technological advancement strengthens cybersecurity operations while preserving the critical thinking skills that human operators bring to complex threat scenarios.