Projets novateurs réalisés

Évaluation de la membrane d’échange de protons durable à base d’hydrocarbures pour applications en pile à combustible

Ce projet MITACS vise à étudier la durabilité des membranes PEMIONTM de l’IONOMR dans les piles à combustible à base de membranes d’échange de protons (PEM) pour des applications automobiles et à explorer leur utilisation comme PEM dans les piles à combustible à base de PEM. Des conditions de test spécifiques et des protocoles à utiliser à l’IONOMR, basés sur les normes industrielles, seront développés et les matériaux seront comparés aux matériaux de pointe actuels afin de prioriser les efforts de développement et d’aider à l’adoption par les clients. L’IONOMR bénéficiera de l’accès à l’équipement spécialisé disponible à l’UBC en plus de son expertise en ce qui concerne les tests et l’optimisation des solutions de pile à combustible PEM fournies par le laboratoire Wilkinson.

International Trade Diversification Opportunities for Canada’s Agri-Food Sector: Lessons from New Zealand

The Canadian economy is built on decades-worth of industrial activity and its links to the natural endowment of the country. For an extended period of time, Canadians have looked to our natural resources for wealth creation, value-added processing, and international market opportunities. On the other end, recent studies and experts have demonstrated concern and warnings against the environmental concerns linked to over-harvesting and over-exploiting Canadian natural resources.

As the Canadian economy continues to rely on natural resource extraction and value-added processes, new markets are emerging outside of the United States of America. For Canada, markets like China are becoming more important as demand continues to soar from Chinese middle-class consumers and an appetite for better and more sustainable foods from Canada.

The research internship, which will take place in collaboration with Dalhousie University’s School for Resource and Environmental Studies and Shanghai Jiao Tong University’s School of Environmental Science and Engineering, will look at the current state of international trading from Canada to China for the agri-food sector, in addition to the future outlook of the Canadian single-use plastic packaging dilemma derived from food products.

Apprentissage par renforcement pour l’analytique sportive prédictive

Our project develops novel machine learning algorithms for interpreting complex, multi-agent scenarios in sports analytics. The collaboration with our industrial partner SPORTLOGiQ will tackle open problems in deep reinforcement learning to build novel capabilities in sports analytics for ice hockey. Deep reinforcement learning is a breakthrough technology with prominent successes in games such as Go (AlphaGo) and Chess (AlphaZero). We will develop fundamental algorithmic advances and apply them to tasks including:
– player evaluation
– event predictions (match outcomes, next action, expected scores)
– recognizing types of players, teams, play sequences, and tactics
– identifying characteristic strengths and weaknesses of players and teams
Montreal-based SPORTLOGiQ uses advanced computer vision to extract information about events from video of sports matches. Their information is more detailed than that provided by any other company or organization. This project will build significant Canadian capacity in sports analytics, support academic research, advance commercialization in the sports industry, and train highly qualified personnel.

Évaluation comparative des méthodes d’apprentissage automatique pour la détection de fraude et l’amélioration de l’interprétabilité du meilleur modèle

Les algorithmes d’apprentissage automatique sont utilisés dans un large éventail d’applications. C’est une branche de l’informatique où le système peut apprendre à partir des données et prendre des décisions. La fraude financière est un risque croissant dans le secteur financier, et il est important de détecter une transaction frauduleuse. Les algorithmes d’apprentissage automatique peuvent être utilisés pour décider si la transaction est frauduleuse ou non. Après que le système ait fait sa prédiction, il est important que les utilisateurs comprennent la raison derrière cette prédiction dans de tels cas. Ce projet de recherche présente un classificateur d’apprentissage automatique pour la détection de fraude qui prédira si la transaction est frauduleuse ou non, ainsi que l’interprétation des prédictions faites par le modèle afin qu’elles soient comprises par les humains.

Une étude Photovoice sur les expériences et besoins de colonisation des hommes immigrants récents dans le centre de l’Alberta

Le Collège Red Deer, avec le Partenariat local pour l’immigration de Red Deer (RDLIP), a reçu des Mitacs Accélération financement pour étudier les expériences de colonisation des hommes immigrants récents dans le centre de l’Alberta. Il y a eu des efforts concertés des gouvernements fédéral, provincial et municipal pour attirer et retenir les nouveaux arrivants dans les villes de taille moyenne et les communautés rurales du Canada; Néanmoins, les expériences globales de colonisation et les besoins des nouveaux arrivants dans ces communautés n’ont pas fait l’objet d’une grande étude empirique; C’est particulièrement vrai chez les hommes immigrants. Cette recherche examinera les expériences de colonisation des hommes immigrants récents dans le centre de l’Alberta afin de combler le déficit de connaissances. L’équipe de recherche partagera les résultats de la recherche avec les membres de la communauté, les organisations de fournisseurs de services de colonisation et les décideurs politiques par le biais d’une exposition Photovoice, d’une présentation en conférence et d’une publication dans une revue académique.

Imaging the Shallow Shear Wave Velocity Structure of the Northern Cascadia Subduction Zone via Compliance Signals Recorded by Ocean-Bottom Seismometers

The Cascadia subduction zone, which extends along a significant portion of western North America from northern California to southern British Columbia, has the potential to produce large devastating earthquakes and tsunamis. Therefore, better understanding the complex structure of the Cascadia subduction zone is paramount to accurately assessing the seismic and tsunamigenic hazard it poses to the western hemisphere. However, this region is difficult to study because it lies entirely offshore. Therefore, this study will image the structure of the subduction zone by using data recorded by instruments deployed on the seafloor, specifically, those belonging to the Canadian NEPTUNE (North-East Pacific Time-series Undersea Networked Experiments) array. One outcome of this research is that it will pave the way for further seismological study utilizing NEPTUNE data. For example, a firm understanding of how seismic energy interacts with the structure beneath NEPTUNE stations is critical to further studies using other imaging techniques. Furthermore this work will showcase the utility of the NEPTUNE arrat for subduction zone research.

Development of Optical Fiber Sensing Devices for Strain Measurementin Rotational Components

Rotational components (e.g., shafts) are key parts of machinery equipment in automotive, aerospace,
energy production, etc industries. These components are continuously monitored for structural
sensing and failure diagnostics by attaching sensing devices on them. This project aims to implement
optical fiber sensors for strain sensing in rotational components. The available technologies for strain
measurement in are prone to electromagnetic noise and disturbance and the methods available to
overcome it have caused high-cost products. Considering the unique features of optical fiber sensors
(e.g., small size, robustness to electromagnetic noise, long-range linearity, etc), they can effectively
replace the electromagnetic devices. This requires addressing the challenges associated with their
implementation which is a goal this internship project. The development of this technology at ATA, will
allow the company to enter into high-tech manufacturing and earn revenue. This will result in the
involvement of Managers, Sales staff, Engineers and production associates for both innovation and
production at ATA.

Une approche intégrée multi-études de cas pour faire progresser les stratégies de marketing en ligne business-to-business (B2B) dans une étude multisectorielle élargie avec 15 entreprises

Le marketing interentreprises (B2B) est considérablement différent du marketing interentreprises (B2C). Les stratégies marketing ne se traduisent pas nécessairement d’un segment à un autre. Cela se voit dans le domaine des médias sociaux et du marketing en ligne, où plusieurs stratégies et outils ont été développés principalement pour le marketing B2C, ce qui conduit de nombreuses entreprises B2B à rejeter les outils de médias sociaux et de marketing en ligne comme étant sans importance pour leur entreprise. Bien que le marketing social et en ligne puisse différer selon un contexte B2B, il ne faut pas le négliger. Une approche stratégique dans le contexte B2B peut stimuler la reconnaissance de la marque, acquérir de nouveaux clients et générer des retours financiers importants pour une entreprise.
Cette proposition vise à mieux comprendre comment le marketing en ligne et les médias sociaux sont mieux appliqués dans un contexte B2B. Nous proposons que l’utilisation d’une approche multi-études de cas offrira aux participants du projet l’occasion de mieux comprendre les outils de médias sociaux/marketing en ligne et de développer des pratiques exemplaires plus pertinentes et applicables pour un environnement d’affaires B2B. La recherche du projet portera sur la couverture et le contenu des médias sociaux gagnés et payants, qui seront analysés à l’aide des outils déterminés par les partenaires du projet (SEM, Hubspot, CRM et/ou Salesforce) afin d’identifier des thèmes et des tendances larges applicables aux besoins marketing B2B du partenaire.

Generation of correlation hypotheses between Adverse Events (AEs) and NamedEntity Recognition (NER) of drugs in social media and scientific journals usinga machine learning approach

Pharmacovigilance (PV) has evolved and grown more complex over the past 5 to 10 years due to increasing data volumes, evolving regulations, influence of emerging markets and the emerging social media and innovative technological advances.
Fast detection of Adverse Drug Reactions (ADRs) could allow the pharmaceutical industry to anticipate and then to control more efficiently eventual risks associated to taking some medications. There is a clear that the social medias data base is important for continuous and automated ADRs surveillance and it may reduce the number of ADR related potential deaths [1].
The scientific literature review is requires for tracking and identification of risk/benefit ratio of drugs and safety issues.
Adverse Drug Reactions (ADRs)

Développement d’un protocole commercial pour différencier les cellules souches pluripotentes humaines en cellules progénitrices du muscle squelettique (hPSC-SMPC) avec un potentiel de transplantation

Recent scientific breakthroughs have led to the development of methods to differentiate human PSCs (hPSCs) into skeletal muscle cells. This has allowed, for the first time, the development of cellular models to study muscle diseases such as Duchenne Muscular Dystrophy and the possibility to utilize these cells for cell therapy applications. However, the reliability, efficiency and prober characterization of cells produced from these differentiation protocols remains a roadblock for their routine utilization by the research community. In this proposal we aim to develop a commercial kit to differentiate hPSCs into skeletal muscle progenitor cells (hPSC-SMPCs) with reproducibility between cell lines with various genetic backgrounds. Building on an existing partnership between STEMCELL Technologies and the Rossi Lab (UBC) this project will perform extensive phenotypic and functional characterisation of hPSC-SMPCs required to validate if these cells represent a therapeutically relevant cell type. We anticipate advancement of this technology will enable researchers to overcome previous roadblocks and facilitate the robust and consistent generation of large numbers of skeletal muscle cells. This new tool will allow researchers to study human skeletal muscle development, model muscle diseases, screen cells for drug discover and investigate the regenerative potential of hPSC-SMPCs in preclinical cell therapy applications.

Guérir le corps, guérir le cerveau : traitements nutritionnels et d’exercice pour les troubles neurodéveloppementaux de l’enfant

In the absence of a clear set of recommendations from doctors for the treatment of

neurodevelopmental disorders (NDD) such as fetal alcohol, cerebral palsy, and autism

spectrum disorders, many parents and children affected by NDDs turn to unproven diet- and

exercise-based treatments. While some research suggests that such interventions can boost

brain functioning, very few studies of these programs look at their effectiveness for NDDs.

This project will produce a long-form piece of investigative journalism seeking to explore the

promises and dangers of relying on diet and exercise to remedy NDD. Sidebar articles with

multimedia components will further examine and provide alternative angles to the use of such

non-drug treatments. Additionally, monthly articles about NDD science and related social and

legal issues will be produced. Work will appear in The Vancouver Sun, in print and online

versions. Ultimately, this work aims to benefit The Sun’s audience by explaining complex

science and ethical concerns in plain language that accurately reflects the facts…..TOBECONTINUED

Curiosity driven reinforcement learning for molecular design

Organic molecules are used in many areas, for example in drugs or solar cells. However, the development of new organic molecules with desired properties is difficult, time-intensive, and requires a lot of trial and error. A new line of research uses artificial intelligence to partially automate this process and make it more reliable. This project aims to further advance the approach. Specifically, the use of curiosity-driven reinforcement learning will be explored. Reinforcement learning is a kind of algorithm, that teaches an artificially intelligent agent to achieve a defined goal, like predicting molecules with desired properties. Curiosity in reinforcement learning generally helps the agent to explore its environment. Here, it has the potential to help agents better explore the chemical space, which can lead to faster development of better molecules that were previously overlooked by human chemists.