Projets novateurs réalisés

Revitalisation socio-territoriale en contexte nordique

Le projet de recherche se déroulera dans la municipalité de Baie-Johan-Beetz, une communauté nordique d’environ 80 habitant·e·s située sur la Côte-Nord du Québec. Cette localité fait face à plusieurs défis communs aux régions éloignées : vieillissement de la population, centralisation des services, rareté des logements et difficultés à attirer de nouveaux résident·e·s. Malgré des efforts importants pour revitaliser le territoire (marketing, développement résidentiel, projets communautaires), la municipalité est limitée par un manque de ressources humaines et une surcharge administrative. Le stage permettra d’appuyer la municipalité de façon concrète, en participant à ses activités quotidiennes tout en menant une analyse des initiatives de revitalisation mises en place. Cela inclut un travail sur les archives, les outils de communication, la sensibilisation citoyenne et l’évaluation des freins rencontrés dans les projets. Le stage offrira un temps de recul et une expertise externe difficile à mobiliser en temps normal. Il visera à renforcer la capacité d’action municipale et à produire des recommandations utiles, tout en valorisant les forces locales. Pour l’organisme partenaire, les bénéfices sont à la fois pratiques (soutien direct) et stratégiques (recul, analyse, transmission des savoirs). Pour la personne stagiaire, c’est une occasion d’approfondir ses compétences en aménagement du territoire, en revitalisation des milieux nordiques et en recherche ancrée dans le terrain.

L2M – Healthcare Systems Performance Optimization using Reinforcement Learning-Enhanced Functional Resonance Analysis Method

Healthcare systems operate as complex socio-technical environments, involving unpredictable interactions among patients, clinicians, technologies, administrators, and institutional policies. These systems are often under significant pressure to improve performance, reduce costs, and enhance patient safety—yet they frequently lack the tools to simulate the consequences of operational decisions before implementation. Traditional approaches like retrospective analysis or static models fail to capture dynamic system behavior and interdependencies, especially when working with limited or sensitive data.

This project addresses that gap by developing a simulation-based decision-support tool that integrates the Functional Resonance Analysis Method (FRAM) with Reinforcement Learning (RL). FRAM helps map functional relationships within healthcare processes, capturing variability in everyday work. RL, in turn, enables the system to learn from simulated experiences by rewarding effective decisions and avoiding poor ones—thereby identifying optimal strategies even in complex and low-data environments.

The aim is to provide hospital administrators, system planners, and policymakers with a user-friendly platform for exploring “what-if” scenarios without disrupting real-world operations. By simulating changes in patient flow, staffing, or resource allocation, the tool can uncover unintended consequences and support more resilient planning.

Through the Lab2Market Validate program, we will conduct structured customer discovery interviews, engage with healthcare stakeholders, and incorporate their feedback to refine the tool’s usability, interface design, and integration requirements. The ultimate goal is to bridge the gap between academic research and real-world application by building a product that is technically robust, practically usable, and commercially viable.

This innovation not only supports more informed healthcare decision-making but also contributes to Canada’s broader priorities around improving public service delivery, fostering applied AI innovation, and enhancing productivity in a resource-constrained health system.

L2M – AR-CAD: An Augmented Reality-based Computed Aided Design Tool

This project will advance the development of AR-CAD, an innovative augmented reality (AR) design tool that allows engineers to create and manipulate 3D models in real-world scale using a headset. Unlike traditional CAD tools that rely on 2D screens and complex interfaces, AR-CAD enables intuitive, hands-on design, helping users visualize and validate parts more quickly. The software exports watertight STL files that are ready for use in 3D printing workflows, significantly reducing design errors, time-to-prototype, and material waste. The partner organization will benefit from early access to a disruptive CAD technology with the potential to improve productivity, reduce costs, and create new market opportunities in rapid prototyping and digital manufacturing.

L2M – Artificial Intelligence Enabled LED Lighting

Cities and lighting companies currently spend weeks and tens of thousands of dollars to create new plastic lenses every time a different streetlight beam is needed or when light pollution must be reduced. Our design is a simple film that sticks to the front of an LED and updates the light patterns quickly—no new lens required, and no delays. We have developed software powered by artificial intelligence that converts any desired beam shape into a printable pattern on this thin film. Once printed, the film is stick to the LED to produce the exact light distribution required. The film weighs less than a paperclip, and can be manufactured in just a few hours. The global exterior LED lighting market is valued at approximately CAD$19? billion. Capturing even 1% of that market represents a CAD?$190? million opportunity. Through Lab2Market, we will identify our initial customer segment—municipalities, stadium operators, or signage companies.

Host-pathogen interaction between a three-dimensional tissue model with a binary cell system and dual species biofilm Under Normoglycemic and Diabetic Conditions

Apical periodontitis is a disease caused by the persistence of microorganisms and their virulence factors in the root canal system. This infection leads to inflammation of the soft tissues and changes in the periapical bone structure. Bacteria such as Fusobacterium nucleatum and Parvimonas micra form biofilms that are resistant to conventional cleaning, making treatment more challenging. Additionally, systemic conditions like diabetes mellitus further aggravate this scenario by compromising the immune response and intensifying inflammation, influencing both the progression and healing of apical periodontitis.
This project will use an in vitro three-dimensional model composed of periodontal ligament fibroblasts and macrophages to simulate the periapical environment. These models will be exposed to bacterial biofilms grown on dentin discs and analyzed under normal and hyperglycemic conditions in terms of cell viability, morphological changes, and interactions with microorganisms.
The aim is to gain a better understanding of the interactions between endodontic infection, periapical inflammation, and systemic conditions such as diabetes mellitus, thereby contributing to the development of more effective and personalized therapies. The use of in vitro three-dimensional tissue models represents an advance in endodontic research, as it enables the analysis of complex cellular interactions that more accurately reflect the real clinical environment.

A novel scaffold for tympanic membrane repair

The ability to hear, and the quality of our hearing, depends on the health of the eardrum. Eardrum perforations due to diseases and accidents can be treated using grafts, such as autologous grafts, allografts and xenografts. These replacements suffer from various limitations such as donor site morbidity, long operation time and healing time, and risk of infection transmission, and more importantly, none of these grafts are able to replicate the complex microanatomy for sound quality reproduced by the native eardrum. In this project, we propose to develop a novel graft with oriented collagen core-shell fibers and growth factors in the core. The oriented fibers mimic the local structure of the native eardrum, producing better sound quality. Furthermore, the delivery of growth factors would
improve healing and promote closure of the eardrum perforation. The successfully developed graft will be commercialized by our industrial partner.

L2M – Augmented Reality-Enhanced Precision Robotic Manipulator Simulation and Training System

The global manufacturing sector is shifting towards high-mix, low-volume (HMLV) production. While large corporations can invest in advanced automation, Canada’s small and medium-sized enterprises (SMEs) face a significant “adoption chasm.” This gap is driven by prohibitive upfront capital costs for robotic systems, a critical lack of in-house programming and maintenance expertise, and the financial risk of investing in technology without a clear return on investment (ROI).

To bridge this chasm, this project proposes the development of an Augmented Reality-Enhanced Precision Robotic Manipulator Simulation and Training System. This innovative platform will function as a low-cost, low-risk “strategic sandbox” for SMEs to embrace Industry 4.0. Leveraging the Unity 3D engine and Microsoft HoloLens 2, the system will provide a high-fidelity digital twin environment where companies can virtually design, prototype, and validate robotic automation processes before committing significant capital.

The system directly dismantles key adoption barriers by enabling clear ROI analysis and mitigating investment risk. It addresses the skills gap with an intuitive, simulation-based training module to rapidly upskill the existing workforce. Furthermore, it enhances operational agility by allowing for the rapid virtual reprogramming of robotic tasks for custom orders and replacing inefficient paper-based workflows with real-time, in-situ digital work instructions.

L2M – Harnessing Adiponectin and Extracellular Vesicle (EV) Biology to Enhance Healthspan

This project explores the commercial potential of a new therapy using adiponectin-enriched extracellular vesicles (Adipo-EVs) to treat age-related cardiometabolic diseases, focusing on stakeholder discovery, market fit, and commercialization planning.

L2M – Road Ally

This project aims to strategically grow Road Ally, a new mobile app that connects drivers facing roadside emergencies with nearby helpers, offering a faster and community-based alternative to traditional services. The intern will conduct market research and develop data-driven strategies to attract more users and helpers, and explore new ways for the app to generate revenue. This work will directly benefit Road Ally by providing a clear roadmap for expanding its reach, increasing its user base, and ensuring its long-term financial stability as it continues to make roads safer and empower local communities across Canada.

L2M – Cross-Modality Translation in Medical Imaging: Bridging Diagnostic Domains

Medical imaging is vital to modern healthcare, yet high costs and limited access slow patients’ paths to
diagnosis. Radiologists often correlate imaging findings across multiple imaging modalities and sequences
to arrive at a specific diagnosis, but each extra scan or sequence increases scanner time, healthcare costs,
wait-list pressure, radiation for ionizing exams, and energy use. An AI image-to-image translator can begin
with the quickest, least-burdensome scan or sequence and digitally create the other needed images, so
patients avoid additional exams or sequences. This saves scanner hours, contrast agents, and energy,
lowers healthcare costs and radiation exposure, and still gives clinicians information required for
diagnosis. The project will survey existing image-to-image translation technologies, gather stakeholder
priorities, and deliver a roadmap that moves synthetic imaging from research into routine care, offering
multiple potential advantages for health systems and patients.

IA Conversationelle pour l’apprentissage du français

Francoflex offre une IA conversationelle qui aide à l’apprentissage du français au travail. Elle aide à l’intégration des travailleurs dont la première langue n’est pas le français.
Pour assurer de l’efficacité du modèle de reconnaissance audio et un bon apprentissage, il faut entraîner un modèle capable de reconnaître des accents forts dans différentes langues natives d’apprenants en français.

L2M – AI-Guided Haptic Robotics for Scalable Surgical Skill Training

Surgical trainees, our primary users in medical schools and teaching hospitals, face a critical barrier: limited access to consistent, expert-guided, hands-on practice. This limitation originates from the constraints of the conventional surgical training model, which depends heavily on the physical presence of expert surgeons. To address this limitation, we are constructing a robot-assisted surgical training platform that enables surgical residents to receive expert-level guidance and correction without constant presence of supervising surgeons. This product addresses core challenges in surgical education, including limited expert availability, duty hour restrictions for surgical residents, reduced hands-on exposure, and the lack of real-time, high-fidelity feedback. The system combines haptic robots with AI, to emulate the motor behavior of expert surgeons. Unlike conventional surgical simulators, which either lack meaningful feedback or depend on surgeons for assessment, our platform provides adaptive, context-aware, real-time guidance through force feedback and motion correction, effectively replicating an expert mentor’s role during hands-on training.