Innovative Projects Realized

An Analytical Model of Thickness and Flatness Estimation for Tailor Rolled Blanks

Rolling is an important process in metal forming. Through the rolling process, metal billets are turned into sheets or strips for use or further processing. Due to worsening global environmental conditions and higher demands for industrial products, lightweight design has become a trend in fields like automotive and aerospace. Traditional rolled metal sheets and strips with uniform thickness are unsuitable for the high demand for complex structural designs. As a result, a new type of sheet with variable thickness has been developed and applied. It’s called Tailor Rolled Blanks (TRB). Research on this new technology is just beginning. This project aims to fill the gap in predicting and analyzing key parameters in the rolling process of TRB. The key parameters will include the thickness and flatness of TRB. The research will advance high-grade sheet manufacturing technology and enhance the levels of industries such as automotive and aerospace in Canada.

Towards a Model-Based Federated Learning Architecture for Intrusion Detection and Mitigation in Small Office and Home Office Networks

This project aims to strengthen RabbitRun Technologies’ (RRT) cybersecurity capabilities in software-defined wide area network (SD-WAN) solutions for small office/home office (SOHO) environments. It addresses the critical need for secure and reliable connectivity in remote and small business settings by integrating federated learning into RRT’s systems. This approach enables RRT routers to collaboratively improve intrusion detection and mitigation while maintaining data privacy and minimizing communication demands.

The key goals of the project are: (1) developing a domain-specific language (DSL) to simplify the mining and specification of security-related network behaviours; (2) advancing state-machine learning techniques to better detect evolving cybersecurity threats; and (3) implementing a federated learning infrastructure for collaborative model refinement across distributed routers.

This initiative bridges academic research and industry innovation, providing the intern with hands-on experience in applied AI, cybersecurity, and model-based software engineering. It also lays the foundation for secure, dynamic networking solutions, benefiting both RRT and the broader Canadian digital economy.

Revolutionizing Asset Management: RFID-Powered Solutions for Asset-Heavy Industries

Scatterlink is a nuanced technology company that delivers advanced inventory and asset management solutions design specifically for asset- heavy industries. Through the use of Radio- Frequency Identification technology (RFID), Scatterlink bridges the gap between physical assets and digital management, providing real-time visibility, tracking, and actionable insights. The asset-heavy industry faces substantial challenges with manual, labor-intensive inventory tracking methods, which result in increased operational costs and productivity setbacks due to asset misplacement and human error. Scatterlink has identified an innovative solution by developing an automated, end-to-end RFID-based inventory and asset management product. The mission of this product is to replace outdated outdated inventory methods with a streamlined digital system that enhances accuracy, reduces operational costs, and optimizes resource allocation. This collaboration with Lambton College seeks to accelerate the development of a minimum viable product (MVP). Lambton College and Scatterlink will develop an integrated mobile and web application using open-source frameworks such as React JavaScript (JS) and Node JS and Express for backend processes tailored to the operational requirements of asset-heavy industries. The successful completion of this project will see the development of a comprehensive digital solution for asset management that will in turn transform the industry. This transformation is essential in mitigating the inefficiencies currently plaguing the industry and fosters a more resilient and competitive economic environment in Canada. This partnership with Lambton College will create approximately 50-75 new high skilled jobs over the next fiver years contributing to workforce development in the technology sector. The anticipated efficiency gains will result in significant cost savings for client companies, positioning Scatterlink to remain competitive in a rapidly evolving global market. Moreover, as this technology offers a proven effectiveness in mining, its applications will extend to various industries, facilitating new market entries and diversifying economic opportunities for Scatterlink.

Innovations in Talent Strategy and Work Culture in Human Resources

The challenge CrucialLogics is currently facing is how to innovate our current processes, workflows, and work culture to support our organization and its clients as we enter the new era of growth and technology. New Innovation in operations (specifically HR) is required as we scale the business and with remote work. This project aims to innovate through identification of gaps, to implement changes in our people operations (hiring operations, training, and performance management). CrucialLogics must adapt and innovate our work culture and approach to ensure that we have the right talent, and that they are supported to achieve success. CrucialLogics is at a turning point as far as its ability to properly support its clients, and to scale to support future ones and HR is required. As different generations and technologies enter both the workforce, as well as that of our customers, its imperative that we continue to innovate and hear ideas and the voices of academia.

Technical Analyst

Technical Analyst to enhance service delivery processes and ensure digitization of key data.

HomeGuard: A Smart Gateway to Protect Smart Home Networks

In this project, the intern student is expected to develop develop a data collection tool that can collect data traffic from IoT devices and label them accordingly. The intern student will then use the tool to collect and label traffic from over 100 IoT devices.

Modèle de mélange avec noyaux pour la classification des données de grande dimension

Les données qu’on rencontre aujourd’hui sont souvent de grande dimension. Avec les données génétiques, les signaux et les images, des méthodes d’analyse qui tiennent compte de la taille des données est plus que jamais nécessaire. Chez Hydro-Québec, une nouvelle méthode de surveillance des équipements électriques a été développée, qui fait appel à la théorie de la communication. Cette méthode a mis en évidence l’utilité de tenir compte du comportement des données aléatoires dans un espace de grande dimension, bien connu en théorie des communications. En adaptant des méthodes de classification existantes pour tenir compte de ce comportement, on pourrait les améliorer. Les méthodes que nous voulons adapter sont basées sur des modèles de mélange. Ils sont très flexible dans le sens qu’ils accommodent des données complexes comme celles provenant de l’expression génique, et elles sont compétitives d’un point de vue du temps de calcul, ce qui est important lorsqu’on traite avec des données volumineuses de grande dimension.

Phase 2 (From Research to Industrialization): Application of Transformer Models to Raw Credit Bureau Files for Improved Credit Risk Modelling Performance

We are developing a new system to help banks and financial institutions better assess credit risk—the chance that a borrower might not repay a loan. Traditional methods use standard data like credit scores, but they often miss valuable information hidden in detailed credit reports. Our project aims to use advanced artificial intelligence, specifically transformer models, to analyze raw credit bureau data more effectively. By doing this, we can create a more accurate and efficient way to evaluate credit risk. This will benefit our partner organization by improving their lending decisions, reducing financial risk, and increasing overall efficiency in their operations.

Functional correlates of central disorders of hypersomnia: a global multi-cohort analysis

The proposed project aims to identify functional brain differences between people with different types of central disorders of hypersomnolence (narcolepsy type 1, narcolepsy type 2, and idiopathic hypersomnia) By combining data from over 1,800 participants from research sites across 21 countries, the project will reveal key brain differences that enhance our understanding of the underlying mechanisms of these disorders. With current diagnostic criteria falling short, this research addresses the urgent need for novel, non-invasive biomarkers that could improve diagnosis and support the development of targeted treatments. The intern will bring valuable expertise in setting up a large-scale MRI analysis to Concordia. She will refine the analysis method and will develop a user-friendly manual that will be used by researchers worldwide. Building on the existing partnership between Concordia and Amsterdam UMC, the project will foster closer collaboration and is expected to lead to high-impact joint publications. Ultimately, it will help position Canada at the forefront of innovative, multi-site neuroimaging research, while strengthening international ties with leading research teams worldwide.

Graph Kolmogorov-Arnold Networks for Anomaly Detection

The proposed project aims to develop a new method for finding unusual patterns in data represented as graphs, such as social networks, communication networks, or financial transactions. We proposed a novel method Graph Kolmogorov-Arnold Networks(G-KANs) to improve how we detect anomalies, which are important for identifying security threats or fraudulent activities. This project will help our participating institutions collaborate and share knowledge in artificial intelligence and its applications. The knowledge gained will benefit both academic research and industry practices, leading to better tools for analyzing data and solving real-world problems..

Terramechanics based Traction Prediction for Lunar Rover by High-Speed Granular Scaling Law

The development of lunar rovers requires research in terramechanics to understand the interactions between wheels and loose terrain. While extensive terramechanics research has been conducted on Earth, few studies have focused on adapting these findings to the Moon’s different gravity environment. This project will focus on the Granular Scaling Law (GSL), a recently utilized mechanical similarity law, and discuss its application to terramechanics. In particular, considering both rover speed and gravity is essential for this application. The host laboratory has world-leading expertise in low-gravity experimental techniques, while the home laboratory is at the forefront of developing experimental facilities for high-speed traversal on lunar terrain. By combining these complementary technologies, this project aims to address the previously unexplored application of the GSL in terramechanics and conduct precise performance predictions for lunar surfaces. If successful, this project is expected to yield significant theoretical insights for both laboratories and contribute greatly to the future development and control of rovers in both countries.

Declining and Stigmatized: An Analysis of French and Canadian Left-Behind Places

The world’s economy is changing at a rapid rate. Resources and population growth are becoming concentrated in a small number of larger cities, while other cities have become “left behind”. The dissatisfaction of living in poorer economic conditions and the resentment toward economically successful cities, has prompted resident political discontent. Left-behind places are also prone to spatial stigmatization, which refers to the way people are devalued and poorly treated due to the places they are associated with, and can affect the way residents and local decision-makers view their surrounding environment. The objectives of this study are to determine the geography of left-behind places in Canada and France, analyze policy interventions in left-behind places to identify how they address left-behindness, and examine the visual transformation of downtown cores in Canadian and French left-behind places. The proposed project will benefit the participating institutions by creating short-term and long-term collaboration opportunities between researchers, as well as opportunities for research dissemination in peer-reviewed journals and international conferences. Furthermore, this study will assess the similarities and differences between left-behind places in Canada and France, informing recommendations which contribute to a better understanding of these places in local, regional, and national level policies.