Projets novateurs réalisés

Large Language Model (LLM): Mobile tools for Social Media Dynamics Creation

In today’s social media landscape, users are constantly creating a variety of content, from texts and images to videos. However, producing high-quality material often requires expertise and resources that many don’t have. This is where our collaboration with PetoLab, an innovative social media app for pet enthusiasts, comes into play. We’re on a mission to simplify content creation for everyone by harnessing the latest developments in Computer Vision and AI language technologies. Our approach is to provide user-friendly tools that operate seamlessly on your phone or computer, leveraging cloud computing for efficiency. Our goal is to transform the way pet stories are shared, making it more engaging and personal for pet lovers everywhere, enabling them to express their passion in unique and captivating ways.

Interferometer Vibration Sensitivity Reduction

General Fusion (GF) is a private company developing a magnetized target fusion scheme to compress a plasma to fusion conditions with a liquid metal liner for the purposes of producing energy. The plasma density is an important parameter that must be measured during a shot to evaluate the performance of the machine. An interferometer is a diagnostic that measures plasma density by sending a laser beam through the plasma and measuring the change in the beam’s phase. GF has a working interferometer on its plasma formation machine, PI3. However, there are issues with vibration that affect the quality of the density measurement. These issues will become even more significant on future plasma compression machines. The objective of this project is to reduce the interferometer vibration error by identifying the source of these issues and modifying the optics and beam path to improve the diagnostic.

Causal Self-Alignment of Large Language Models

“THIS IS A GENERIC TEXT PUT IN PLACE AS THERE WAS NO PROJECT OVERVIEW”

Paysage sonore Aviqtuuq

La recherche proposée vise à soutenir l’observation de l’environnement marin dans la région Aviqtuuq autour de Taloyoak, Nunavut, en lien avec l’initiative en cours de développement d’une aire protégée et de conservation autochtone (APCA) incluant un programme d’intendance du territoire Aviqtuuq par la population locale. À partir d’enregistrements acoustiques collectés par trois hydrophones déployés sur des mouillages sous-marins lors de la saison estivale 2023, l’étudiante fournira une caractérisation du paysage sous-marin des trois sites étudiés. L’analyse permettra de quantifier les niveaux de bruits ambiants observés durant la saison estivale, à identifier et quantifier la contribution anthropique via le trafic maritime, et à décrire l’utilisation de l’habitat et le comportement acoustique des espèces marines sonifères, notamment la population de bélugas.

Electrical property measurements of breast tissues: image reconstruction and links to clinical information

Traditional breast monitoring via mammograms is avoided by 37% of eligible women for multiple reasons, including discomfort and inconvenience. To overcome this, there’s a rising interest in advocating microwave
imaging as a safe, affordable, and portable alternative. This encourages early and regular monitoring, especially for younger women, with the added benefit of accessibility in rural areas.
In this proposal, we use scans collected with a microwave imaging system to identify biomarkers in the images. This aids in categorizing breast density, a crucial breast cancer risk factor. The system’s capability to detect changes in related to the presence of tumors enables the identification of cancer biomarkers. We also aim to monitor treatment progress, providing a secure tool for regular use during chemotherapy to track tumor changes. The reconstruction algorithms for generating breast dielectric properties and deriving biomarkers are beneficial for our industrial partner.

Northern Capacity & Development Project – Sambaa K’e

Sambaa K’e is a remote indigenous community in the Dehcho region of the Northwest Territories. It has approximately 90 residents, and most of the year is reachable only by small aircraft. Building on an established research partnership with the Sambaa K’e First Nation and Sambaa K’e Development Corp, we propose to send interns to live in the community and lead a series of innovation and development projects to benefit the community and improve the economic prospects of the area. Our research has identified one major barrier to improving food security and economic goals in Northern indigenous communities is limited access to skilled labour capacity. These interns will gain valuable life experience, learn about other cultures, and help the community achieve it’s economic development goals and bring food sovereignty a step closer.

Assessing Climate Resilience and Mitigation in Manitoba: Aligning Municipal Strategies with Sustainable Development Goals

The Rural Development Institute (RDI) and Eco-West Canada are studying 135 Manitoba municipalities, focusing on accessing the state of their climate mitigation and adaptation/resiliency plans. This initiative aims to evaluate existing climate action strategies, identify gaps, and align them with funding criteria under the Mobilizing Municipalities program. Through research and collaboration, the project aims to develop an engagement and research plan that outlines outreach methods, communication techniques, and evaluation criteria. A detailed record of current climate plans will be compiled. The project will provide a detailed analysis of Manitoba’s municipal climate change plans, including insights, best practices, and strategies for enhancing local Manitoba climate mitigation and resiliency plans. The findings from this research will inform future policies and programs in Manitoba and other regions. By emphasizing collaboration and knowledge sharing, the project allows policymakers to engage with partners, gain valuable perspectives, and shape data-driven insights for policy engagement. There will be a preliminary and final report, assessing how well climate plans align with the criteria of the Mobilizing Municipalities program. The project will also include relevant Sustainable Development Goals (SDG) in an analysis framework to identify the current state of those change indicators for rural Manitoba. For example, Canada’s SDG indicator 13.1.1 -Proportion of municipal organizations who factored climate change adaptation into their decision-making process will be used as an analysis indicator in this research project. The results will be compared to the Mobilizing Municipalities program’s criteria to identify potential action areas. RDI and Eco-West Canada aim to use this study to analyze and promote climate-focused activities in Manitoba municipalities, advancing towards a sustainable future.

Graph-Driven Strategic Intelligence: Innovations in Forecasting, Delisting, and Marketing Optimization – Part 2

THIS IS A GENERIC TEXT PUT IN PLACE AS THERE WAS NO PROJECT OVERVIEW

Fuel cell diagnostics through integrating model-based technique with electrochemical impedance spectroscopy (EIS)

The wider adoption of fuel cells has been hindered by durability and reliability issues. This research alliance (consisting of Ballard and SFU) has made significant recent advances on developing novel diagnostics techniques to detect the inception of hydrogen leak faults in fuel cells. We aim to continue to build on these advances through this grant.
This research alliance has been working towards truly online diagnostics techniques for detection life limiting faults in fuel cells. Previous successful projects had focused on diagnosing leak faults. This proposal will focus on monitoring the hydration of (level of water within) the membrane used in the fuel cell as loss of hydration is a major degradation mechanism. To this end, this project will develop an enhanced model for fuel cell hydration and simulate different fault levels to generated a fault profile detector.

Squeezing water from rocks: Moon ore and mineral process prototype for water production

For human space exploration to be successful, technologies to produce breathable air, drinking water, and fuel are needed to sustain life. Particularly on the Moon, there are a variety of minerals and ore in different regions on the Moon’s surface that can be extracted and chemically processed to produce air, water, and fuel. The objective of this project is to test and scale up a novel process, invented by CSMC, for producing water and oxygen from minerals that are prepared to simulate the composition of the Moon’s soil. Once this is accomplished in a scalable way, it will be possible for humankind to establish a Moon colony for resource extraction and further space exploration. The benefit to the partner will be to have a completed prototype that is designed to maximize the water produced from the Moon’s soil.

Autonomous Cable Seal Application and Removal

The two major objectives of the umbrella grant are to 1) empower the partner organization to improve their digital capacity and gain insight into their data to support business growth and innovation. 2) Provide students with work-integrated learning experiences that expose the students to advanced data analytical techniques, data mining and machine learning. DICE focuses on a companies Data, Networking and Analysis (DNA) which is often unique to each company and involves solutions that are as unique as the DNA in humans.

This project is projected to be a part of a larger project 3 years in length and will require interns to grow their knowledge in multiple different areas of technology including Machine Learning (ML), IoT and modern software development frameworks.

Rayhawk is providing autonomous railcar loading technologies for granular loading facilities. This project entails enhancing Rayhawk’s existing vision system to work efficiently and without issues as Rayhawk scales it’s customer base and railcar support. Rayhawk has built a machine learning pipeline that has currently satisfied our first production unit within an inside facility focusing on one particular railcar, but as we scale we need to support multiple railcars as well as support them in all weather and all condition with a railcar either stationary or in continuous motion. Data research and analysis is needed to identify steps required to provide a scalable vision system across all of these variables.

This project will focus on integrating LiDAR and other sensors such as cameras and infrared sensors in conjunction with a robotic assembly that will be used to interact with railcars during the loading processes.

Graph-Driven Strategic Intelligence: Innovations in Forecasting, Delisting, and Marketing Optimization – Part 1

THIS IS A GENERIC TEXT PUT IN PLACE AS THERE WAS NO PROJECT OVERVIEW