Innovative Projects Realized

Super resolution for MRI scans

Brain MRI scans are a critical component in the diagnosis of neurodegenerative disorders. However, there is a wide diversity in terms of the image quality and resolution obtained from different MRI scanner. In particular, it is common to find coarse resolution MRI scans (e.g. every axial slice is 3-5 mm thick), which limit the type of anatomical analysis that can be performed. The goal of this project is to develop and validate the performance of state-of-the-art super-resolution methods in 3D MRI scans, which generate high resolution MRI scans from low resolution scans.

Low data drug modeling

The project aims to facilitate the research and development of new drugs by exploring Machine Learning methodology useful for both the generation of new molecules and the prediction of molecule properties. Doing so will involve training deep learning models on a large number of small, heterogeneous datasets, with the objective of transferring learned representations quickly when faced with a new drug-discovery or drug optimization objectives. The trained models will be used for the purposes of predicting molecular properties of new drugs and generating novel molecules with high likelihood of satisfying certain properties. The multi-objective nature of designing new molecules satisfying competing objectives will be approached using techniques from Reinforcement Learning.

Altering Plant Microbiomes for Flavour and Nutrition

The goal of this project is to use naturally occurring bacterial partners to improve the flavour and nutritional properties of plants grown in hydroponic and aquaponics systems. This study will investigate ability of plant associated bacteria to alter the metabolic profile of select vegetables and leafy greens. Vertical farming is an increasingly popular solution for the production of plant produce year-round at a local level. However, it involves the growth of plants in engineered systems without natural soils. Soils are inhabited by tens of thousands of species, some of which move into plant tissues and contribute to their natural nutritional and flavour profiles. Our aim is to match produce plant metabolisms to those of naturally occuring bacterial associates and to test the ability of these partners to enhance the quality of food produced in vertical farms.

Development of Copper Precursors for Atomic Layer Deposition

Microelectronic fabrication needs a method to deposit very thin films of copper in very accurate patterns to interconnect the microelectronic devices on a chip. Atomic layer deposition (ALD) is a method used in microelectronic manufacturing that could do this, but a suitable copper process has yet to be identified.

This internship is to help continue to develop a promising copper deposition process for ALD. Using a volatile compound containing copper, it has been shown that copper metal can be deposited at reasonably low temperature using a plasma of hydrogen gas. The intern on this project will continue to develop the copper chemistry and the deposition process, as well as help the partner organisation (GreenCentre Canada) to construct and file a patent application for this intellectual property.

Cognitive and Computationally Intelligent Algorithms for the Detection of Cyber Threats

The rapid and widespread advancement of cyber-threats within the past few years has had a profound impact on virtually everyone, from ordinary people to governments to local organizations. This has caused cyber-security to be considered a global challenge, which is now requiring innovative solutions, such as incorporating human cognition based methods into the software algorithms to detect malicious activities of adversaries. This is because, the cyber-security industry is heavily dependent on the knowledge, and analysis and investigation skills of analysists in the detection of cyber-threats. The analysists have the ability to coalesce and examine in their minds large stacks and disparate sources of data, spread across small and large temporal and spatial windows, and compare their observations with previous known attacks to decide if the dataset objects under scrutiny represent an attack. TO BE CONT’D

Mitigating heat stress effects on lactation via nutrition in dairy cows

Heat stress, defined as temperature and humidity above the threshold of animal comfort, is experienced in Canadian dairy farms each year despite the improvements in cooling conditions and results in important reductions in production of milk and of milk components. Such reductions represent lower production efficiency and therefore higher environmental impact per unit of product and significant losses in profitability for producers and for the processing industry. Importantly, dietary interventions focused on the mitigation of production loses could be easily implemented in dairy farms across Canada and result in improved production efficiency and profitability for the dairy sector. In the present proposal we set out to test the effects of increasing the supply of nutrients capable of mitigating heat stress-related production loses (select lipids, amino acids, minerals and vitamins). TO BE CONT’D

Modélisation du marché d’espaces publicitaires en ligne avec des séries temporelles

Les espaces publicitaires que l’on retrouve sur la plupart des page web font partie d’un marché similaire à celui des actions. Les espaces publicitaire sont mis à l’enchère sur une plateforme d’échanges électroniques et les publicistes misent sur ces espaces afin d’y afficher du contenu ou de les revendre. Dans ce marché relativement nouveau, les stratégies d’achat et de revente sont souvent déterminées par un agent qui surveille le marché et qui adapte ses stratégies en fonction de ses observations.
L’objectif de ce projet est d’utiliser les connaissances existantes en mathématiques financières pour modéliser ce marché et de développer des outils quantitatifs d’aide à la décision. Ces outils permettront d’optimiser les stratégies pour rendre les entreprises plus profitables et, en conséquence, rendre ce marché plus efficient.

Improving the Performance and Convergence Rate of Transformer-Based Language Models

The pre-trained Bi-directional Encoder Representation from Transformers (BERT) model had proven to be a milestone in the field of Neural Machine Translation, achieving new state-of-the-art performances on many tasks in the field of Natural Language Processing. Despite its success, it has been noticed that there are still a lot of room for improvement, both in terms of training efficiency and structural design. The proposed research project would explore the detailed design decision of BERT on many levels, and optimize them wherever possible. The expected result would be an improved language model that achieves higher performance on NLP tasks while using less computational resources.

The (re)regulation of cannabis: a comparative analysis between the Netherlands and Canada

Research has shown that the traditional field of crime and cannabis control is disrupted. Hence, several jurisdictions are moving towards decriminalization and even legalization of cannabis. The Netherlands was one of the first countries to adopt such an alternative approach by allowing small retail of recreational cannabis. In October 2018, Canada decided to legalize cannabis. Yet, both countries are still struggling to establish a well-functioning cannabis policy that focuses as much on public health and human rights as it does on crime control.
This research aims to investigate the regulation of cannabis in the Netherlands and Canada, examine the area of conflict with the international drug treaties and provide an explanation for the regulation of cannabis in the Netherlands and Canada. TO BE CONT’D

Evaluation of monitoring data for predictive maintenance of energy production assets

Hydro-Québec has data acquisition systems for a multitude of sensors, some of which have been installed since almost 20 years in its electrical generation equipment (turbine-generator units – TGU). The collected data is primarily used to ensure that the information is adequate in the event of an equipment breakdown or for specific behavioral studies. Data from monitoring systems are little used in routine maintenance management activities, often due to lack of time and adequate and/or effective analysis methods.
Equipment maintenance is an important part of Hydro-Québec’s equipment management activities. The creation and maintenance of a surveillance system is a major investment for the company. With the development of machine learning analysis approaches, the goal is to provide operators with a clearer view of real-time asset status and predictions about their potential for use.

Feature extraction from 2D and 3D images of flocs

Flocculation is a chemical treatment to improve the settling of fine particulate suspensions in the food, chemical, agricultural, and mineral processing industries. The size, shape, and density of the flocs influences their mechanical strength and settling rate; which are critically related to the efficiency of solid-liquid separation processes. The objective of this research is to test the theoretical predictions of 3D structure from 2D images of flocs by a direct comparison of 3D predictions and 3D measurements of floc structure. The broader context of the proposed work is to improve our understanding of the mechanisms of densification and thickening in difficult-to-treat mineral suspensions so that improved solid-liquid separation (SLS) treatment methods and technology may be designed. From an environmental standpoint, improved SLS processes result in decreased fresh water needs, a smaller ecological footprint, and a decreased likelihood of tailings spills.

Ground-Based Remotely Piloted Aerial Vehicle (RPAV) Tracking System

Drone Delivery Canada (DDC) designs and operates high performance Remotely Piloted Aerial Systems (RPAS) to deliver payloads between depots and warehouses. The DDC engineering department is looking to design and deploy a ground-based system to track and point at the Remotely Piloted Aerial Vehicle (RPAV) during flight in real-time. However, DDC’s RPAS must be able to operate in remote areas making the use of communication technology infrastructure difficult due to the need to be able to have communication between the RPAS and ground control station (GCS) over long distances without the use of communication relay nodes or heavier, more powerful communication modules on the RPAV. To solve this challenge, advanced communication equipment such as high-gain antenna(s) will be needed in addition to novel antenna tracking algorithms for the system to be interfaced with the GCS to receive telemetry data from the RPAV. The end product is a robotic tracking system which applies positional feedback data from the RPAV (such as altitude and GPS location) to a control system to dynamically point an antenna at the RPAV throughout flight. Review of existing literature on tracking systems and required infrastructure/resources would be performed to guide the design process. TO BE CON’T