Projets novateurs réalisés

Additively manufacturing SMR used fuel cladding – simulation of material radiation damage for leak testing

Due to chemical reactivity of components in new small modular reactor (SMR) fuel types, e.g., sodium, the federal nuclear regulator has identified spent SMR fuel as an R&D priority area. With multiple sodium fast reactors, including TerraPower’s Natrium reactor, and ARC Clean Technology Canada’s (ARC) ARC-100 reactor, planned for deployment in Canada within the next decade (ARC-100 to be deployed in New Brunswick by 2030), research surrounding the acceptance of spent nuclear fuel into a deep geological repository (DGR) is critical for the successful commercialization of SMRs in Canada.
In this project, additive manufacturing (3D printing) of the cladding alloy proposed for the ARC-100 reactor will allow controlled defect creation, simulating neutron bombardment during reactor operation. In the worst-case “what-if” scenario, where a DGR containing sodium-cooled fuel is flooded, water would interact with sodium that has diffused into the defects, making these necessary for accurate bench-scale experiments studying water-sodium interactions.

Effects of Metal Ions on Biological Membranes and Cellular Systems

Heavy metal toxicity has been a cause for public health concern for decades. There are a variety of industries that both produce heavy metals as a byproduct, or use them directly in consumer products or practical applications. In particular, cadmium (Cd) is one such toxic metal that is known to induce negative effects not only on cellular systems, but to organs as a whole. Cd is a known carcinogen, and can impart cellular toxicity through a number of biochemical pathways. While some of these pathways, such as DNA damage, production of reactive oxygen species (ROS), and inhibition of apoptosis are well categorized; the effects of Cd at the level of cellular membranes are less well known. Cellular and organellar membranes not only provide key structural platforms, but also serve as signalling platforms and act as interfaces between cells and between organelles. To understand the mechanisms by which Cd may induce damage to these crucial structures is key to completing a full outlook on the potential dangers associated with Cd toxicity.

Luxembourg Research Visit

Deep Reinforcement Learning (DRL) for robotic grasping has been actively studied in recent years. Each DRL needs a reward function to interact with its environment and figure out how desirable or suitable the actions it takes in each state are. The reward formulation of DRL is usually a linear summation of the reward components, which is inefficient to learn the multi-objective priorities. Hierarchical reward methods have been proposed to enable a robot to learn multi-objective tasks such as achieving autonomy or human-like merging actions for driving. The formulation of the reward hierarchies contains logical or weighted connections. Logical connections are strict constraints, where the higher-level hierarchy must be learned before the lower-level hierarchy. Weighted connections are soft constraints, where the higher-level hierarchy and lower-level hierarchy are learned together with a weighted. For this project, what I have in my mind is that instead of using a single coded reward like linear summation, use an extended hierarchical reward function to tackle the problem in a way that even includes gripper actions as a decision variable and can learn multi-objective tasks. In my work, these multi-objective tasks could be divided into approaching, attempting to grasp, grasping phase, and stabilizing.

Évaluation d’impact des programmes d’accompagnement des entrepreneurs collectifs : Cas du programme SISMIC

Au Québec, l’entrepreneuriat collectif compte 11200 entreprises collectives avec un chiffre d’affaires cumulé de 47,8 milliards de dollars et emploie près de 220.000 personnes (Institut de la statistique du Québec, 2019). Le Chantier de l’économie sociale a lancé différents programmes de soutien aux entrepreneurs collectifs. Dans ce sillage, le présent projet de recherche vise à évaluer l’impact d’un de ces programmes d’accompagnement pilote sous le nom de “SISMIC”. Ce dernier se présente comme un réseau d’incubateurs d’entreprises collectives offerts aux jeunes partout sur le territoire québécois et auprès des communautés autochtones. Ainsi, cette étude vise à évaluer l’apport de cet accompagnement sur le développement des projets d’économie sociale et des entrepreneurs bénéficiaires. Pour ce faire, nous avons opté pour une approche qualitative à travers l’organisation d’entretiens semi-directif avec les entrepreneurs collectifs ayant bénéficié de cet accompagnement au cours des dernières années.

Adaptive and Emotionally Aware Chatbots

This research focuses on the development of machine learning and natural language processing techniques for automated textual dialog systems called chatbots. More specifically, we will develop algorithms to allow chatbots to adapt to new topics and concepts in open ended domains. We will also develop algorithms to recognize the emotions of users through their text messages and select emotionally adequate responses. The outcome will be emotionally aware chatbots that can adapt to new domains. This will address an important need for automated dialog tools that allow organizations to engage with users in a personalized fashion via text messaging platforms. For instance, a music band or a film producer may launch a chatbot that impersonates the lead singer or main character to interact with fans, answer questions, announce new releases, etc. Kik Interactive, a leader in text messaging, will directly benefit from this research as it seeks to market such communication tools to its partners.

Optimiser le spectre lumineux afin de favoriser le contrôle biologique dans les productions végétales en milieu contrôlé

L’alimentation durable est une composante essentielle de l’économie visant à un développement apte à répondre
aux besoins du présent, sans compromettre la capacité des générations futures à répondre aux leurs. Pour ce
faire, l’alimentation durable préconise la consommation de produits locaux et, en ce qui concerne les produits
maraîchers, cultivés dans le respect de l’environnement. En régions urbaines ou péri-urbaines, la culture en
milieux contrôlés (serres ou bâtiments), rendue possible grâce à l’éclairage DEL, constitue une avenue des plus
intéressantes.
L’objet de la présente étude sera la manipulation du spectre lumineux afin de promouvoir l’action d’un prédateur
naturel, et ainsi d’optimiser la saine croissance de végétaux sans l’utilisation de pesticides. Positionner le Canada
en tant que leader de la culture biologique et durable en milieux contrôlés aura un effet bénéfique tant sur
l’économie que sur la santé de l’ensemble des Canadiens.

Expressive Speech-to-Face

This project aims to enable controllable generation of expressive, speech-driven facial animation in video games and multimedia. Current methods use procedural tools to animate mouth movements given speech clips, but they lack realism. This research proposes to develop a new approach for controllable speech-driven facial animations, combining the realism of recent mesh deformation methods with the ability to control the animation process. Using machine learning and a dataset of expressive facial performances, the research will focus on learning speaking styles, and the artistic manipulation of these styles, to achieve more nuanced and convincing animations. The anticipated outcome is improved facial animation quality in video games, benefiting the partner organization by providing a tool that enhances character expressiveness and overall gaming experience.

Development of high-performance earthquake resilient tall buildings

With rapid population growth worldwide, major cities around the world are seeing an upsurge in high-rise building construction. At the same time, there is an increased awareness about the importance of designing buildings so that they will have limited damage due to strong earthquake shaking. An innovative structural system is proposed that utilizes an outrigger in combination with a conventional reinforced concrete (RC) core-wall system. To improve the seismic performance of the structure, innovative fuses are added to the outriggers to dissipate the sudden surge of the earthquake energy. In addition, replaceable coupling beams are used to ensure the damaged portions of the structure can be easily inspected and quickly repaired after strong earthquake shaking. In this research, innovative structural fuses will be developed through international research collaboration between the Earthquake Engineering Research Facility Canada (EERF) and Tongji University (headquarter of ILEE center). Full-scale components will be systematically designed and tested at the EERF facility in Canada, while large-scale shaking table tests will be conducted at Tongji University to validate the overall seismic performance of the proposed system. The proposed research project will significantly contribute to the development of high-performance earthquake resilient tall buildings.

AI-Driven Carotid Ultrasound Advancements: Redefining Stroke Risk Assessment in 3D

Stroke, a leading global cause of death and disability, often results from the rupture of dangerous fatty deposits, known as plaques, in the neck arteries. The current method of assessing stroke risk using 2D ultrasound is limited by operator dependency and lacks a 3D structure assessment, leading to misdiagnoses and re-scans. Our project aims to overcome these challenges by developing AI-powered algorithms that enhance the accuracy of 2D ultrasound diagnostics, improving stroke risk assessments and treatment decisions. This innovative approach will position Canada as a leader in healthcare innovation, contributing to advancements in AI disciplines and promoting accessible and inclusive healthcare solutions. The partner organization, Sonaro, will benefit by gaining access to high-quality data, ensuring the accuracy of the developed models, and establishing valuable network connections within the academic institution

See One, Do One, Teach One…Value One: Integration of Artificial Intelligence in Encounter-Based Evaluation Application to Help Teaching Be More Valued

The goal of this project is to incorporate AI to complement and upgrade existing functions in myTE. We plan to invest in 3 features: thematic analysis, data interaction and web analytics. In our current version, myTE compiles all comments in learners’ feedback and presents them in their original form to the users. Using AI, the feedback can be summarized by theme, spellchecked, and paraphrased – which enhances our protection to learners’ anonymity and facilitates ease for user’s CV curation. In terms of data interaction, AI would make use of data available, such as ratings in each type of encounters, as well as teacher recordings, to make meaningful inferences on teaching quality and teaching recommendations. Finally, web analytics support continuous user interface improvement in the backend and analysis on user demographics (e.g., gender, ranking) and how they affect evaluations. We will apply machine learning to look at which functions users are capitalizing on and enhance on those, while removing or switching up some of the less capitalized features.

A Scoping Review of Training Curricula for Youth Peer Supporters in Early Psychosis Programs

This scoping review will explore the research base of the intersections of peer support training and implementation. The partner organization will benefit from this project as it will form the foundation of a longer-term goal of developing a training module for early psychosis peer support workers.

Synthesis of Novel Coumarin Derivatives with Anti-Neuroinflammatory Activity