Innovative Projects Realized

ML-based link prediction algorithm for RWA

Artificial Intelligence (AI) and Machine Learning (ML) are key technologies in the development of a new recommendation framework that is able to identify the association between users and items with greater accuracy. This link prediction algorithm can be used in economics, marketing, networking, and social media. It would decrease the rate of mispredictions and would allow businesses to target their customers properly and increase their profit margin. The aim of this project is to develop a ML-based link prediction algorithm for the Refresh Wellness App (RWA) to discover the associations in its mental health resources system and provide the best resource recommendations for users. The performance of the recommender system will be evaluated in terms of accuracy. Moreover, the reliability of the developed model will be validated with case studies.

Resources Module Strategy

Research is currently underway to develop evidence-based resources that will populate the resource section. In the meantime, a robust infrastructure must be established within the app to adequately support an extensive selection of resources for a variety of users and environments. Because not all of the resources will be relevant to all users, a dynamic solution is needed so that only relevant resources are stored on users’ devices. Therefore, this project aims to establish a strategy for this dynamic feature delivery system. In order to test the implementation, this project also includes the development of a resource called the “break timer”, in which employees are invited by their workplace to take breaks at specified times. By the end, the RWA platform will include a dynamic resource delivery system, as well as a “break timer” to demonstrate the new capabilities.

RWA Market Launch and Brand Strategy

While the Refresh Wellness App (RWA) is in the research and development phase, Refresh has begun to build a marketing strategy that will aid in the launch and distribution of the app to ensure a positive impact on society as a whole. While the RWA will eventually be available to the public, Refresh plans to first launch the platform in a workplace setting. While workplace stress isn’t a new concept, workplace burnout is on the rise since 2020 – driving a major resignation rate as a result. “More than two-thirds (70%) of employees said they would leave their current job for a different one that offered comprehensive resources, benefits, support or policies intended to reduce burnout.” [8] Because of this goal, a dynamic solution is needed so that relevant users can be targeted in a way that is both meaningful to them and aligns with the values and mission of the RWA.

This project aims to develop a marketing strategy based on a fulsome understanding of a) the key target demographics throughout the different phases of the Law of Diffusion of Innovation, b) the ideal marketing platforms that will align with the mental wellness landscape and RWA’s dedication to ethical data practices, and c) key messaging factors related to different geographic and cultural target markets and their views on the mental health and wellness landscape.

Demand Forecasting Analytics Prototype

The partner organization, Granularity LTD, helps businesses use demand forecasting to capitalize on market trends. At the start of the pandemic, half of all grocery stores in the U.S. were out of toilet paper despite manufacturing being operational. The inconsistencies of demand in the COVID-19 pandemic awoke business leaders to the realization that their demand forecasting capabilities can’t keep up. After speaking to subject matter experts in the space in over two dozen market interviews, including demand planning executives for large retailers and manufacturers, there is a clear gap in understanding end customer behaviour and how it impacts demand. This is a burning need that we are aiming to solve. We want to build a solution that specializes in conveying end-customer behaviour and insights to businesses. The key to understanding this behaviour is data. Our proposed project will infuse demand forecasting with machine learning and big data; incorporating social media sentiment, google search trends and other datasets to enhance demand prediction. Now, a grocery store using our solution would have the demand forecasting capability to pre-emptively see the increased demand for toilet paper — allowing the business to drive millions in potential revenue.

Correlation of Coronary and Myocardial Strain Profiles for Cardiac Resynchronization Therapy Year Two

Throughout my Doctoral Studies I have used various approaches to evaluate and/or predict the behaviour of cardiovascular structures in the setting of acquired disease. These have been focussed on two key methodologies: insilico testing aimed at predicting functional responses to therapeutic intervention, such as mechano-electric Finite Element Method (FEM) modelling of the left atrium strains and stresses in response to hypertension, and fluiddynamic FEM modelling of abdominal aortic aneurysms. This has allowed me to gain knowledge about selection of appropriate medically relevant approaches to simulation with respect to different biological structures, and about proper boundary condition setting in order to accurately set the correct environment within which organ-level simulations are run. Finally, I have learnt how to optimize such computational approaches in order to be able to run strain and stress mathematical modeling within clinically acceptable turnover times. Through such education, we can obtain proper prediction of organ response to different pharmaceutical, electrical or mechanical surgical therapeutic approaches, considering the different physics involved by the problem (fluid-dynamics, mechanical tissue behaviour and electrophysiology).

Assessing the geohazards of Mount Meager, BC, Canada: Insights to advance green energy production on volcanoes

Climate change is leading to the retreat of glaciers and more extreme weather events that greatly increase landslide hazards on Canadian mountains as seen by the recent catastrophic events and fatalities in British Columbia. BC also hosts dormant volcanoes which are important areas for development of geothermal energy yet are vulnerable as they consist of hydrothermally altered and weakened rock. By integrating satellite remote sensing with ground-based monitoring, and geological and geotechnical information, this project aims to help the Meager Creek Development Corporation better assess and effectively mitigate the slope stability risk within their leasehold. Importantly, we aim to develop a framework for semi-automated landslide monitoring of potentially unstable slopes around their geothermal energy and green hydrogen production facility on Mount Meager. The hope is that the results of this research can be applied more broadly on volcanoes and alpine regions throughout Western Canada.

Stage Stratégie & Développement d’affaires Données & RSE

En 2022, Axionable souhaite accélérer la pénétration du marché nord-américain et se positionner comme le leader de la donnée et de l’IA au service du développement durable en proposant de nouvelles offres novatrices pour contribuer à la transition écoresponsable en Amérique du Nord. Axionable mettra à contribution ses nombreux partenariats dans le milieu universitaire (IVADO, CRIM, UdeM, OVBIA) et industriels au Québec et en Ontario afin de développer et commercialiser ces nouvelles offres.
Dans le cadre de notre stratégie 2022-2023, nous avons par exemple priorisé le développement de solutions numériques permettant la mesure et la limitation des émissions de GES. La finalité de ces solutions a par conception un impact durable et positif tant sur l’environnement que la société.
Le stage de stratégie et développement d’affaire données & RSE que nous proposons s’inscrit dans la phase d’opérationnalisation de notre stratégie 2022. Gaspard aura l’opportunité unique d’intégrer un environnement de travail en constante ébullition, bienveillant, international dans lequel il mettra à contribution ses compétences acquises pendant sa Maitrise en Gestion au HEC afin de contribuer à relever les défis suivants :
– Opérationnalisation et ajustement de stratégie corporative et développement d’affaire (réalisation d’études de marché, exécution de plan stratégique, Planification et suivi des performances…)
– Compréhension du marché de l’innovation en valorisation des données et IA pour le Développement durable.
Implication de la compagnie : Gaspard travaillera au quotidien avec le VP Exécutif Canada
Implication du superviseur académique : Il supervisera Gaspard pendant la durée de son stage dans l’atteinte de ses objectifs.

Hybrid Precoding/Combining (HPC) and Self-Interference Cancellation (SIC) for Full-Duplex (FD) Massive-Multiple-Input Multiple-Output (mMIMO) Wireless Communications Systems

The rapid advancement of wireless communications has drastically transformed the world over the last few decades. The next-generation wireless communication systems are expected to support a wide variety of requirements for diverse applications (e.g., the Internet of Things (IoT), industrial IoT (IIoT), autonomous driving, healthcare, virtual/augmented reality). Meeting such stringent requirements will bring significant challenges and require the exploration of new enabling technologies. The objective of this project is to develop advanced and innovative physical layer techniques to satisfy this urgent demand. More specifically, we will focus on the development of intelligent signal processing techniques for full-duplex massive multiple-input multiple-output (FDmMIMO) systems. Toward this end, the combination of artificial intelligent (AI) and convex optimization will be the main tool to analyze and address the design problems.

Design, implementation and feasibility testing of a mindfulness meditation intervention to reduce neuropathic pain among individuals with a spinal cord injury

Tens of thousands of Canadians living with a Spinal Cord Injury (SCI) suffer from debilitating neuropathic pain. This project, in partnership with Spinal Cord Injury Ontario (SCIO), will seek to understand and test the potential for mindfulness mediation, to decrease pain and increase quality of life. Over 36 months, Robert Buren (the intern), in collaboration with UBC and SCIO, will work together to learn from relevant research that has already been completed, survey the experiences and needs of the SCIO member community, and finally, create and conduct a mindfulness mediation program with a small group of the SCIO community to test if mindfulness mediation can help them reduce pain and improve their quality of life. Throughout this project, the learnings and findings will be shared with the SCIO community, to ideally help everyone in Ontario, Canada, and abroad, living with SCI, to reduce their pain and improve their lives.

Nbwaakaawining binjibaamgad Gkendmaawziwin Wisdom from Knowledge: Documenting and Sharing the Indigenous Biocultural Richness of the Greenbelt

Plenty Canada has partnered with the University of Guelph to launch a biocultural knowledge and mapping project to begin restoring Indigenous knowledge, visibility, and character to the Greenbelt as an important Indigenous cultural landscape in Ontario. Our proposed work builds upon the success of a pilot project that we launched with the Canadian Commission for United Nations Educational, Scientific, and Cultural Organization (CCUNESCO) to document and safeguard important Indigenous heritage resources along the Niagara Escarpment. We are now planning to expand our research and mapping project to the entire Greenbelt region. The intern will assist Plenty Canada to gather Indigenous knowledge and scientific data and resources to establish a repository of information that will aid public education and serve as the foundation for enhancing awareness of the importance of preserving natural world ecosystems and green spaces to safeguard Indigenous flora through the braiding of Western science and Indigenous Knowledge systems.

Metabolomic Biomarkers of Traumatic Brain Injury

Clinical diagnosis and rehabilitation of mild traumatic brain injury (mTBI) relies on patientreported symptoms and on limited availability of quantitative assessment tools. Also, there is no clear standard of care and lot of disparity in the rehabilitation protocols used in the clinic. The industrial partner APEXK Inc. has developed novel patented technology-based solutions to enhance precision of assessment and neurorehabilitation for mTBI patients both in athletes and non-athletic clinical populations through a standardized protocol (Class II Health Canada approved). Current approaches by APEXK use oculomotor and visual-cognitive modalities to deliver personalized treatment. The addition of a quantitative metabolomic signature would enhance precision in mTBI diagnosis, monitoring rehabilitation progress and recovery. The academic team has identified and patented a systems biology-based metabolomics signature with high predictive accuracy that will address this urgent medical need. The novel combination of APEXK’s unique neurorehabilitation methods combined with our patented novel biomarker technology will provide an innovative strategy to match quantitative physiological genomicbased biomarkers to a Health-Canada approved data-driven technology used for concussion rehabilitation.

Sensors for Monitoring Septic Tanks

We are developing an automated monitoring system to optimize the frequency of septic tank pumps.
Septic tanks are used in ~87% of rural communities across Canada; current best practice for
maintenance pumping is neither effective nor efficient as it is based on time rather than use. A
reliable monitoring system can reduce cost by minimizing the frequency of pumping (more efficient)
and avoid environmental damage associated with overflow (more effective). There are three layers
in a tank: floating scum, liquid effluent, and bottom sludge. Both scum and sludge build up over time;
if the tank is not pumped, solids will clog the outflow pipes in the connected leach field. In this
project, we will first perform field testing to determine which layer is most effective for determining
when a tank needs to be emptied. Then we will compare different sensor technologies in a lab test
setup. Thereafter, we will prototype an enclosure and a data logging system for the most effective
sensing technology. Finally we will field test the sensor prototype.