Innovative Projects Realized

L2M – Lucy DNA Biotechnology

LucyDNA introduces a cutting-edge genetic testing platform specializing in personalized healthcare solutions. Leveraging polygenic risk scores provides comprehensive genetic risk assessments for chronic diseases, offering tailored disease prevention, screening, and treatment recommendations. Targeting hospitals, mature individuals, and health organizations, LucyDNA enhances healthcare planning with precise insights, ensuring optimized health outcomes and precision medicine.

L2M – Biomimetic biointerfaces for high throughput in vitro modeling of gut-pathogen interactions

The recognition of gut microbiota, its interplay with human tissues and cells, and the signaling pathways they could generate in concrete with together, specially once imposed to gastrointestinal medications, encouraged researchers designing and establishing bioengineered platforms, enable to serve as a discovery tool for developing of microbiome-related therapeutics, monitoring medications combating detrimental effects of antimicrobial resistance (AMR) pathogenic species infections or carcinogens on human epithelial cells, or checking probiotics and nutraceuticals benefiting human health. We are introducing a bio-inspired intestinal villus mimicking structure composed of natural extracellular matrix, that primarily adheres human intestinal epithelial cells without any significant cytotoxicity, stimulating tight epithelium formation, cell polarization, and differentiation into strains capable to accumulate dense mucosal barrier potentially functioning as a site for microbial colonization, controlling their invasion into underlaying intestinal cells and verotoxin generation. The established balanced microenvironment for cell-bacteria co-culture brings us the opportunity to monitor the crosstalk among these species in details. In light of this innovative bioengineered material mimicking human body conditions and physiology, the need for in vivo testing for pre-clinical assessments of new drugs and medications is minimized. Traditional methods based on in vivo models, whose translatability to human’s body condition is frequently challenged, are time-consuming and often require skills in animal surgery. The outcomes of this innovation facilitate the rapid screening of gastrointestinal medications, consequently can significantly reduce the burden on the Canadian and global healthcare systems.

Cost-Effective and Intelligent Monitoring of Large-Scale Software Systems

This research project aims to tackle the challenge of ensuring the reliability of large-scale software systems, which are crucial for various applications such as online services and data processing. The project focuses on developing cost-effective and intelligent methods for monitoring these systems to anticipate and prevent runtime incidents, such as crashes or errors, which can disrupt operations and lead to costly downtimes.

Throughout the project, the goals will be achieved through a systematic approach involving data examination, statistical analysis, and machine learning techniques. By analyzing monitoring data and building predictive models, the research will successfully identify patterns indicative of potential runtime incidents, empowering proactive measures to mitigate risks and enhance system performance.

The project provides numerous benefits to all involved parties. For the student, it offers practical experience, skill development, mentorship, networking opportunities, and contributions to academia. The academic supervisors benefit from enriched collaboration, reputation enhancement, knowledge exchange, and potential publication opportunities. Additionally, the host institution gains from the student’s contributions to ongoing research efforts and potential future collaborations.

Improving AI by Learning Directly from Human Preferences

Our project aims to improve how AI models learn from human feedback. Current methods assume human preferences can be reduced to a single “reward” value, but research shows this isn’t always true. We will investigate if an AI algorithm can learn from human preferences without relying on rewards. If possible, we’ll design and test algorithms which learn without reward. If not, we’ll explore the cases where rewards are necessary. This research will promote AI systems that are better aligned with human preferences, benefiting the scientific community and industry sectors like healthcare and education. The insights gained from this project will advance Canada’s AI research and reinforce both Mila and Stanford’s global leadership in AI.

Satellite image enhancement & crop classification

Our research project focuses on monitoring crops using satellite images. To address the problem, we leverage a fast-growing field of graph signal processing (GSP), which expands upon traditional signal processing techniques such as Fourier transform and wavelets to accommodate the graph domain. Specifically, we propose to unroll a designed graph-based algorithm into an interpretable feed-forward neural net for end-to-end parameter tuning, in order to enhance satellite images and classify field crops. The primary benefit for the public is the development of a robust and efficient model for crop monitoring. The model is resilient against the known covariate shift problem, which occurs when the distribution of input data differs significantly from the training and test datasets. Unlike conventional “black box” deep learning models, our model requires much less data for training to train significantly fewer network parameters, enabling an operator to save resources and time. With this technology, an operator can swiftly and accurately assess crop health and productivity, leading to informed decision-making and optimized agricultural practices.

Growing Sustainable Tourism in Edmonton’s Chinatown

Edmonton’s Chinatown has long served as a vibrant cultural, social and commercial hub for generations of Asian immigrants and local residents. Yet it is characterized by a complex blend of social and economic challenges. This proposed project, “Growing a Sustainable Tourism in Edmonton’s Chinatown,” is designed to accelerate efforts to revitalize Edmonton’s Chinatown through planning and developing recreation and tourism products and experiences. This project involves collaborative development of a Tourism Strategy for the neighbourhood that will strengthen community and working bonds amongst stakeholders, document diverse perspectives on the best opportunities for Chinatown to explore regarding tourism investment, and inspire collective action amongst government, NGO, researchers and community members to accelerate roll-out of new tourism offerings.

Digging deep and looking up: Applying Canada’s mining experience to the moon

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

Les approches sensibles aux traumas : de la théorie à la pratique

Ce projet s’intéresse au déploiement d’approches sensibles aux traumas en contexte d’intervention pour des
populations en situation de précarité sociale. La notion de traumatisme complexe réfère à l’exposition,
particulièrement pendant l’enfance et l’adolescence, à des traumatismes interpersonnels répétés, et aux réactions
comportementales associées à ces événements négatifs dans de nombreux domaines du fonctionnement
(Greeson et al., 2011). En réponse aux taux élevés et aux conséquences négatives des traumatismes complexes
sur le fonctionnement, le bien-être et les comportements des personnes, des approches d’intervention dites
« sensibles aux traumas » ont été développées dans les dernières décennies. L’approche sensible aux traumas
se démarque puisqu’elle sous-entend qu’une organisation entière doit devenir « sensible aux traumas », et ce,
par une transformation de ses pratiques organisationnelles (Purtle, 2018). L’objectif du projet vise évaluer
l’adéquation entre l’offre de services et pratiques du partenaire et les approches sensibles aux traumas. Pour
réponse à cet objectif, la stagiaire réalisera une recension des données scientifiques et des entrevues avec
différents groupes (intervenant, presonnes premières concernées, etc.). Les retombées du projet visent à outiller
la stagiaire dans les méthodologies de recherche et de permettre au partenaire de documenter ses pratiques.

Accelerating electronic structure simulations on quantum computers using complete active space approximations

Computational chemistry is a powerful tool in the development of new pharmaceuticals, materials, and batteries. With accurate and fast computational methods, one can predict the physical and chemical properties of a molecule without having to prepare it in a laboratory and perform experiments, which can yield dramatic time and cost savings. However, the chemistry of molecules is determined by the laws of quantum mechanics, which are challenging to simulate on even the largest supercomputers that exist today. One resolution of this problem is to use so-called quantum computers which can perform computations that encode these challenging quantum properties intrinsically at the hardware level. As building quantum computers is a challenging engineering problem, the current quantum computers that exist are very small, and can only perform computations for a short time before errors build up and scramble the results. For chemistry, this motivates the question of how to develop more compact representations of the target molecule one intends to simulate so that a smaller quantum computer can be used to study its chemistry. In this project, we will develop and benchmark various methods to generate compact representations of chemical systems, and determine which are best for quantum computing.

Self-supervised Representation Learning via Self-Evolvable Random Projections

Self-supervised representation learning (SSRL) has advanced considerably by exploiting the transformation invariance assumption under artificially designed data augmentations. While augmentation-based SSRL algorithms push the boundaries of performance in computer vision and natural language processing, they are often not directly applicable to other data modalities, and can conflict with application specific data augmentation constraints. This project aims to propose an SSRL approach that can be applied to any data modality and network architecture because it does not rely on augmentations or masking.

L2M- GreenLith- Using Cellulose Beads to capture Lithium to recycle batteries

Every technology today, from phones to EVs, rely heavily on lithium-ion batteries, which account for more than 90% of the world grid market. While these batteries make us independent of fossil fuels, we must remember that lithium is a finite resource, hence making it crucial for maintaining a secure Lithium supply and ensuring proper disposal of Li-ion batteries. Therefore, with our core values in innovation, sustainability and cost-effectiveness and guided by UNSDGs 9,12, and 13, we present to you GreenLith- specialized cellulose beads acting like tiny sponges, that would selectively capture Li ions from shredded lithium batteries

Monster House Publishing & UNB Arts 4000 Internship – Marketing and Publicity Coordinator

Working with staff and industry partners, the intern will build and implement a cross-platform marketing strategy that utilizes existing and potential promotional structures within Monster House Publishing. The intern will work within the company, learning the various aspects of the publishing space while finding new and creative ways to increase the sales and visibility of Monster House both locally, nationally, and internationally