Projets novateurs réalisés

Sustaining an Early Detection Program for Childhood Cancers in Cameroon: A Qualitative Study

Globally, sustaining evidence-based innovations (EBIs) in public health and health care settings remains a critical challenge, with up to 60% of EBIs failing to maintain the delivery of their core components and intended health benefits after initial adoption and implementation. This is especially relevant in low- and middle-income countries (LMICs), where EBIs to strengthen childhood cancer and other health programmes are increasingly being implemented without sufficient research attention to their long-term sustainment. To address this gap, the proposed project aims to examine how an early detection program for childhood cancers is being sustained across multiple health districts in Cameroon, combining quantitative and qualitative methods to understand the health system factors that support or hinder its sustained delivery and impact. By working closely with local health leaders and pediatric oncology clinicians, the study will generate practical, context-specific insights to inform how early cancer detection programs can be strengthened and sustained in Cameroon and similar low-resource settings, ultimately improving timely diagnosis and treatment for large populations of children with cancer.

Quantification des textures pétrographiques de l’or et du graphite

Cette demande de subvention propose de développer une plateforme d’analyse d’image basée sur la vision par ordinateur destinée à la pétrographie appliquée à l’exploration minière canadienne. Le projet vise à créer deux modules?: (1) un algorithme automatisé de mesure de la taille des paillettes de graphite, offrant une évaluation rapide, économique et hautement reproductible des paramètres influençant le prix du graphite?; (2) un outil d’analyse de la texture de surface des grains d’or provenant de sédiments glaciaires, permettant d’inférer l’origine et le degré de transport des particules, indicateurs clés pour identifier des gisements potentiels. En combinant ces capacités, nous améliorerons la précision et la rapidité des évaluations géologiques, réduirons les coûts d’exploration et renforcerons la compétitivité du secteur minier canadien. La subvention demandée financera le développement d’un algorithme, la validation sur jeux de données réels et le transfert de technologie vers le partenaire industriel.

Hybrid Physico-Data Framework for Predictive Modeling of PEM Fuel Cells

This project aims to develop computationally efficient and reliable computer models to help design longer-lasting hydrogen fuel cells. We combine our physical understanding of how these devices work with modern artificial intelligence tools to predict performance losses and lifetime under realistic operating conditions. During the internship, I will work with Prof. Marc Secanell’s Energy Systems Design Laboratory at the University of Alberta and my home institute at Forschungszentrum Jülich/RWTH Aachen to link their advanced electrochemical models with new machine-learning approaches. The collaboration will create practical modelling tools that can be used by researchers and industry partners, strengthen long-term cooperation between Canada and Germany in clean energy, and provide specialized training at the interface of engineering, materials science, and AI.

Sahati Fiyadi/Ma santé, mon pouvoir: la santé sexuelle et reproductive des femmes migrantes au Maroc

En raison de sa proximité géographique avec l’Europe, le Maroc est devenu un lieu important de transit, voire de résidence prolongée ou finale de personnes migrantes et réfugiées. Au cours de leur parcours migratoire, ainsi que dans les pays d’installation, ces personnes sont souvent confrontées à de multiples obstacles pour accéder à des soins de santé appropriés et adaptés. À ce sujet, les études démontrent que les filles et les femmes migrantes sont plus à risque de subir de la violence, notamment sexuelle, et elles sont également plus souvent exposées à des facteurs de stress et à des conflits liés à leur rôle social. En conséquence, elles souffrent généralement de problèmes de santé sexuelle et reproductive (SSR) spécifiques. Pourtant, nous avons encore peu de connaissances sur leurs attentes et leurs besoins dans ce domaine. Les objectifs de cette recherche sont donc d’identifier les attentes, les besoins et les déterminants contextuels influençant l’accès des femmes migrantes aux services de santé sexuelle et reproductive au Maroc. Les résultats du projet permettront de renforcer ou d’adapter les politiques publiques au Maroc, ainsi que de proposer des réformes des cadres juridiques applicables.

6G-SEER: Exploring Scalable, Energy-Efficient and Resilient AI-Native 6G

Wireless communication networks play a pivotal role in modern society, connecting people, devices, and digital services. While 5G networks are now widely deployed, emerging applications such as smart healthcare, autonomous transportation, intelligent infrastructure, and immersive digital experiences will require much more capable systems. Sixth-generation (6G) networks are expected to meet these needs by providing faster, more reliable, and more intelligent connectivity. However, 6G will also face major challenges, including the need to support massive numbers of devices, operate under strict energy constraints, and remain reliable in highly dynamic and complex environments. These challenges motivate the integration of artificial intelligence (AI) to enable networks that can learn, adapt, and operate with minimal human intervention.

The 6G-SEER project addresses these challenges by focusing on three essential objectives: scalability, energy efficiency, and resilience. Scalability ensures that future networks can grow and support dense populations of users and devices. Energy efficiency aims to reduce power consumption and support sustainable network operation with low-fingerprint impact. Resilience focuses on maintaining reliable communication despite disruptions, failures, or changing conditions. By exploring AI-native approaches, 6G-SEER seeks to enable self-organizing, adaptive, and robust 6G networks that can meet future societal and technological needs.

Towards scalable quantum chemistry in quantum computers through localized domain ansatze for unitary coupled cluster.

This project will develop a more efficient way to run quantum-computing simulations of molecules, an important step toward faster discovery of new materials and catalysts. Today, many quantum chemistry methods on quantum computers become too expensive as molecules grow, requiring large numbers of circuit operations and measurements. We will adapt a proven “local” strategy from classical quantum chemistry (which focuses computation only on the most important electron interactions) to build smaller, shallower quantum circuits, and we will test mappings that better preserve chemical locality on quantum hardware. The work will be carried out through a research internship, sending a PhD candidate from McMaster University to the University of Augsburg, combining McMaster’s strengths in scalable quantum algorithms and measurement-efficient methods with Augsburg’s expertise in quantum chemistry circuit design and the Tequila quantum software ecosystem. The outcomes from this project include an open-source implementation of our localized method on quantum devices, aiming to use it in the near future to train machine learning models for reducing quantum resource utilization, bringing to Canada highly qualified personnel prepared to develop scalable quantum algorithms, adapt highly specific quantum chemistry workflows, and perform high-impact international research.

Seismic Response of Deep Foundations in Liquefiable Ground: A Numerical Investigation

Pile foundations support critical infrastructure such as bridges, ports, and energy facilities, particularly in earthquake-prone regions. During strong earthquakes, saturated soils can undergo liquefaction, a process in which soil temporarily loses strength due to increased pore water pressure. This can impose large vertical and lateral forces on pile foundations, increasing the risk of structural damage or failure.

Current design standards acknowledge liquefaction risks but do not provide clear or unified methods to accurately predict pile behaviour under these conditions. As a result, engineers often rely on simplified assumptions that may underestimate structural demands or lead to overly conservative designs.

This project aims to improve the understanding and prediction of pile foundation performance in liquefiable soils using advanced three-dimensional numerical modelling. The research will simulate key mechanisms such as pore water pressure buildup, soil–pile interaction, downdrag, and lateral ground movement during earthquakes.

A central component of the project is model validation using high-quality field data from the 2012 Emilia earthquake in Italy. By comparing numerical results with observed field performance, the project will develop more reliable modelling approaches and practical insights for engineering design. The knowledge gained will contribute to safer, more resilient infrastructure and strengthen seismic engineering practices in Canada.

Microfluidic phage display platform for ligand evolution in 3D cellular constructs and engineered phage therapeutics delivery

This project aims to develop a microfluidic device that can identify cancer-targeting molecules capable of penetrating and retaining tumor tissue. Traditional laboratory methods identify the targeting molecule using 2D cell cultures or solely protein as the target model. However, neither of these models can accurately represent the structure and conditions of real tumor tissue, leading to the failure of most candidates in clinical trials. To minimize reagent use and make the device low-cost and portable, I use consumer-level 3D printing to fabricate the mold for the microfluidic chip. 3D cancer cell models that reproduce tumor physiology were trapped in the chip to screen for molecules that bind and penetrate the models under realistic conditions. The mature device and screening workflow will be transferred to the host institution, where advanced biological validation can be conducted using complementary expertise and facilities. This collaboration allows the Canadian institution to advance microfluidic engineering and screening technology development, while enabling the host institution to apply the platform to downstream validation and broader experimental applications. Together, the project establishes a transferable, reusable screening framework that supports international collaboration and enhances cancer-targeting research.

Privacy-Aware XAI for Opinion Mining in Healthcare and E-Governance

Digital healthcare and e-governance platforms have become central to service delivery, citizen engagement, and policy implementation. These platforms generate vast volumes of unstructured textual data in the form of patient feedback, citizen complaints, online reviews, and public opinions. Analysing this data provides valuable insights into service quality, institutional trust, and public perception. Opinion mining and sentiment analysis techniques enable organisations to systematically extract such insights. While these techniques have been successfully applied in commercial domains, their use in sensitive public domains raises significant concerns related to data privacy, transparency, and ethical accountability. Existing systems often prioritise predictive accuracy while neglecting explainability and privacy protection, limiting their suitability for healthcare and public-sector decision-making. This project proposes a privacy-aware and explainable artificial intelligence framework for opinion mining, specifically tailored for digital healthcare and e-governance systems. By integrating sentiment analysis with privacy-preserving mechanisms and explainable AI techniques, the project aims to deliver trustworthy, interpretable, and ethically aligned analytical insights that support responsible decision-making.

L2M – Market Entry and Human Factors Strategy for an Edge AI-Enabled Care Platform: Converting Pilots into Repeatable Sales

NewGen Health is developing a connected “Smart Urine Cup” and supporting patient + clinic software to modernize urine screening in maternal health. Today, urine dipstick testing is still largely manual and subjective, which can create inconsistent results and adds operational burden in high-volume prenatal care. NewGen’s platform is designed to make urine screening more standardized and connected, enabling clinics to capture quantified results through a controlled workflow and support longitudinal monitoring in both in-clinic and hybrid/virtual care settings.

This project will help NewGen convert recent clinical traction, into a repeatable, scalable clinic offering. We will translate clinician and pilot feedback into clear workflow requirements, usability improvements, and standardized onboarding/training materials so clinics can deploy the solution with minimal disruption. In parallel, we will develop the commercialization foundation needed for sustainable growth: a clinic deployment playbook, a refined ideal customer profile, procurement-ready materials, and a pricing/packaging approach that supports recurring revenue.

The expected benefit to NewGen Health is a practical path from pilot execution to multi-site adoption. By reducing implementation friction and improving sales-cycle repeatability, the project will strengthen NewGen’s ability to close additional clinics, shorten time-to-deployment, and build a sustainable business model while maintaining a rigorous Human Factors/HCI lens aligned with AIDL.

Using Photogrammetry as a Storytelling Tool

At the Broken Parts is an experimental animated short film that explores memory, displacement, and childhood resilience using emerging 3D technologies. The project uses photogrammetry and 3D scanning to capture real-world spaces and objects, which are then transformed into animated environments inside a game engine. These environments are combined with sound and visual effects to create a poetic, non-verbal story told from a child’s perspective.

The intern(s) will help research, test, and develop this workflow by assisting with 3D scanning, animation, scene assembly, and visual experimentation. Through hands-on production, they will learn how real-world data can be turned into expressive animation using contemporary digital tools.

For the partner organization, this research supports the development of new creative technology pipelines that can be reused across future film, installation, and immersive media projects. It strengthens internal capacity in photogrammetry-based animation, reduces production risk for future work, and contributes to innovation at the intersection of storytelling, digital media, and emerging visualization technologies.

Scalable Markerless Motion Capture: Building a Biomechanical Database for Data- Driven Coaching and Exercise Prescription

This project will use video cameras and computer vision to measure how people move during common exercises and golf swings. Instead of placing markers on the body in a lab, the system will estimate joint positions directly from regular video, making movement analysis cheaper and easier to use in real-world settings. The intern will record people performing different exercises, process the videos to calculate how their bodies move, and identify which movements are most effective or may increase injury risk. The partner organization, BIOVEKTR, will use these data to improve their “digital twin” models of the human body, helping them provide more accurate, personalized feedback about movement and exercise to users in fitness, sport, and health applications.