Projets novateurs réalisés

La recherche par pertinence avec l’utilisation d’un moteur d’intérêts à représentation visuelle

Avec les moteurs de recherche actuels, l’utilisateur a accès à de plus en plus d’information : la quantité d’information à analyser augmente de jour en jour et l’utilisateur ne peut consulter toute cette information parce que tout simplement il n’en a pas le temps : après avoir regardé les résultats de la 1ière ou 2ième page de Google le reste est ignoré. Comment filtrer efficacement cette information pour la rendre pertinente pour l’utilisateur? Une solution est de filtrer les résultats en fonction des intérêts multiples des utilisateurs mais selon la classification DEWEY une liste complète d’intérêts de référence peut aller jusqu’à + 10,000 intérêts. Comment donc, en quelques manipulations de l’utilisateur, identifier ses différents intérêts avec exactitude et par ordre d’importance ? Puisqu’une photo vaut 1,000 mots la solution explorée pour solutionner cette problématique est le développement d’un algorithme permettant la présentation dynamique de photos préalablement catégorisées et l’analyse itérative des sélections individuelles pour déterminer les intérêts personnels de chacun en lui présentant les photos suivantes en fonction de ses choix précédents, suivi d’une retro-action des participants pour valider la sélection des intérêts identifiés par notre algorithme.

Générer des images de haute qualité des produits de chez Portes Milette à l’aide de méthode traditionnelles et d’apprentissage machine.

Ce projet vise à générer des images de haute qualité, des différentes portes de chez Portes Milette. L’entreprise permet à ses clients de choisir une multitude de paramètres afin de personnaliser ses modèles de portes. Les clients ont accès à des millions de combinaisons de paramètres possibles. La visualisation de ces portes personnalisées par le client avant même leur conception est le défi que doivent relever les 2 stagiaires Mitacs dans le cadre de ce projet. Le tout sera réalisé à l’aide de méthodes d’intelligence artificielle de même qu’à l’aide de méthodes d’ingénierie traditionnelle. Pour l’entreprise Porte Milette, ce nouveau système permettra de consolider l’une des forces de l’entreprise, soit d’offrir des produits hautement personnalisés qui seront maintenant visibles sur leur site web.

Data-driven prediction of project issues using AI

A considerable percentage of BNC projects are impacted by stoppages, re-adjustments and delays. Such issues are often communicated too late to prevent significant losses or facilitate the reallocation of funding to maximize value creation for the bank and its clients. The majority of delivery metrics (risk, statuses, stakes) that help to anticipate strategic issues and take action are based on the experience of project managers. The democratization of project management platforms such as JIRA now allows the bank to access more precise project data and track project activity. On-time identification of even a small percentage of such projects can facilitate the reallocation of funds and save millions of dollars in funding.
The goal of this project is to use ML to model the probability of project stoppage and delay. The initiative will collect actionable data and key performance metrics from the project management toolset at BNC and utilize such indicators to predict the overrun of the budget and/or schedule of features and alert the right upstream stakeholders to take action.

Leveraging the Canadian Longitudinal Study on Aging to Support Healthy Aging Programming in the Community-Based Seniors’ Services Sector

Transportation problems are one of the most common barriers, after health problems, that prevent older Canadians from participating in more social, recreational, or group activities. United Way Healthy Aging is working closely with the community-based seniors’ services sector to bring solutions to seniors’ transportation. Through the Better at Home program, United Way Healthy Aging supports a network of over 80 agencies across the province to deliver seniors’ transportation services, typically in the form of seniors’ shuttles, buses, and volunteer driver programs. One challenge to advancing transportation programming in this sector, and the work of the non-profit sector more generally, is limited capacity to locate, collect, and analyze data which can be used to quantify the scale of problems, inform where efforts should be targeted, and advocate for change. In this project, I will bridge this data gap by determining how data from the Canadian Longitudinal Study on Aging – the largest and longest multi-year study on aging – can be used to strengthen the sector’s ability to advocate for and deliver transportation programming.

RWA Content Marketing Development and Launch – Refresh

The Refresh Wellness Platform intends to offer a novel approach to address ongoing mental health and wellness challenges worldwide. While the RWA is in the research and development phase, they have 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. This project aims to develop the aforementioned marketing strategy based on a fulsome understanding of the key target demographics, the ideal marketing platforms that will align with the mental wellness landscape and the key messaging factors related to different geographic and cultural target markets. By the end of this project the RWA will have a Content Marketing Strategy for 1, 3 and 5 year growth, as well as have enacted their brand launch.

Plan stratégique pour la commercialisation et l’exportation de produits liés à l’érable – QC 476

Vendre à l’année et exporter nos gammes de produits.
L’activité principale de la Ferme Cantin & Laroche inc. est agricole et acéricole. Tandis que L’Érable au Fil du Temps inc. est chargée de produire et de commercialiser les produits transformés à base du sirop 100% pur d’érable. Nos projets d’innovation se diversifient et s’étendent au niveau de l’automatisation de nos procédés, de la logistique, la structure, les activités, les finances et le financement.
Nous voulons participer et contribuer à développer l’acériculture, devenue aujourd’hui une activité économique à part entière, puisant sa source à même une tradition ancrée depuis des siècles dans la culture québécoise. Pour y parvenir nous voulons faire découvrir l’Érable partout ! Grâce à nos cornets !

Randomized Controlled Trial of a Behavior Change Intervention to Increase Aerobic and Resistance Exercise and Quality of Life in Older Prostate and Breast Cancer Survivors: The OutPACE Trial

Age is a key factor in cancer burden, with 28% of new cancer cases and 22% of cancer deaths occurring in individuals between the ages of 60-69. Despite this, little research has examined cancer survivorship in older adults. This is unfortunate, as older cancer survivors experience poorer quality of life, functional status, and general health compared with individuals with no cancer history. Older cancer survivors also are at an increased risk of developing recurring and secondary cancers, as well as other chronic diseases (e.g., cardiovascular). Research has established that exercise and physical activity can improve cancer survivors fatigue levels, strength, quality of life, well-being, health and physical functioning while preventing cancer recurrence. However, few cancer survivors are meeting physical activity guidelines for aerobic or resistance training. The purpose will be to explore the utility of a behaviour change intervention on aerobic and resistance exercise in older prostate and breast cancer survivors; and explore effects of the intervention on quality of life, fatigue, and physical functioning.

Système de mise à jour de la cartographie routière basé sur les données géospatiales ouvertes

E-SMART conçoit un écosystème de produits qui offre aux flottes de camions une façon innovante d’assurer le respect des limites de vitesse par les conducteurs de camion. Dans le cadre de ce projet, la compagnie souhaite développer une technologie permettant d’utiliser la cartographie publique, par exemple OpenStreetMap (OSM) et de faire la mise en correspondance (« map-matching ») entre la trajectoire du véhicule et le segment de route correspondant, afin de savoir quelle limite de vitesse est rattachée. Ce développement technologique permettrait à E-SMART de mettre en œuvre des améliorations qui ne sont pas possibles en utilisant la solution commerciale utilisée par la compagnie.

Les buts à terme sont les suivants :
? Améliorer la précision des limites de vitesse retournées par le système, notamment en améliorant la mise en correspondance temps-réel ou map-matching.
? Diminuer la quantité de données transférée lors de la mise à jour des cartes.

Evaluation of Therapeutic Natural Killer cells

Cellular immunotherapy has yielded numerous clinical benefits in treating cancer. While many successes involve CAR T cells, unfortunately this potent anticancer cell therapy can cause serious and even fatal complications. Another immune cell, the natural killer cell (NK), may lead to effective treatment without unwanted side-effects. Furthermore, the use of stem cell-derived NKs provides additional advantages, including the ease of gene editing and ready availability of an off-the-shelf therapeutic. We have comprehensively characterized several potential NK products and demonstrated that they are highly cytotoxic against a variety of cancer cells. The intern will extend these studies to engineered NK cells that can specifically kill cancer cells expressing certain proteins.
Ankarys Therapeutics is a Canadian early stage biotech company that uses gene editing and stem cell technology along with scientific and clinical expertise to develop proprietary cell therapy products consisting of safe, powerful, engineered stem cell-derived NK cells.

Investigation and Development of an Automated Testing Platform for Digitally Implemented Clinical Guidelines in the Health Level (HL7) Fast Healthcare Interoperability Resources (FHIR) Standard

The project partner, JuniperCDS (Juniper), is a digital health company that is developing a care and treatment planning (software) application called JuniperCDS, which is a clinical decision support software solution that enables clinicians to create meaningful, person-centric care plans by delivering in-context recommendations based on the latest clinical guidelines at the point of care. A major challenge with clinical guideline digital implementation is testing the guideline to ensure it is complete and meets HealthLevel7 (HL7) Fast Healthcare Interoperability Resources (FHIR) standards, Currently, testing a guideline is a highly manual process with minimal existing tools that are difficult to use. With this project, Juniper would like to develop a digital testing tool that automates a major portion of the testing and provides significant time savings. The tool will evolve Juniper’s apps. Externally, the testing tool will be a global public good for testing and validating digitized guidelines.

Sales Micro-Patterns: ML-driven detection of deviant Store/SKU sales patterns

For almost a century, Canadian Tire has been offering products and services to help Canadians with the jobs and joys of life in Canada. Today, with a network of over 1,700 retail locations, Canadian Tire is dedicated to improving products and services offered to the community they serve. With this project, the company aims to examine the patterns of sales among its stores to identify anomalous product sales patterns. The objective is to design and develop a system capable of generating high-confidence automated notification to the stores. Sales patterns will be investigated to identify products deviating from normal patterns. The stores would be notified of the results and incentivized to investigate the flagged Stock Keeping Units (SKUs) for in-store issues, such as missing inventory or tags.

The impacts of biomass-burning aerosols on air quality, weather and renewable energy in British Columbia

In the last few years, wildfires’ have become more severe and frequent in British Columbia, Canada. In addition to the direct dangers of flames, threatening wildlife and human society, as well as vital infrastructures, wildfire plumes laden with Biomass-burning (BB) aerosols can adversely affect air quality, weather, and renewable energy generation on a large scale. BB aerosols can disrupt solar radiation, thus disturbing the surface energy budget, atmospheric stability, and cloud evolution.
The proposed international research will study the plume’s dynamics and its interplay with the atmosphere through an exhaustive computational campaign to shed more light on the mentioned impacts and enhance the Canadian’s analyzing and forecasting capability to address them.
For that purpose, an intern will be trained in ICON (ICOsahedral Nonhydrostatic), a high-resolution numerical weather forecasting model, at KIT, particularly its Aerosol and Reactive Trace gases(ART) module. The intern will then set up the model on SFU’s supercomputer upon his return and exercise the acquired knowledge on analyzing large-scale forest fires emissions.