Projets novateurs réalisés

Geochemical analysis of low-sulphidation epithermal deposits on the La Victoria property in the Ancash Department of the Republic of Peru

This MITACS research project will focus on analytical techniques related to determining the gold geochemistry of the La Victoria property in the western Peruvian Andes. The La Victoria property is found along a major gold belt that is host to some world class deposits such as Yanacocha, Lagunas Norte, La Arena and Shahuindo. Based on initial exploration techniques, two major mineralized areas (San Markito and Rufina) have been identified as important for gold exploration. The next stage of exploration is to examine the geochemistry of surface samples and drill cores to determine the mineralogical and geochemical signatures of the gold mineralization and to put these into a consistent model. Essentially, the mineralogical and geochemical data obtained in this study will be beneficial for continued exploration of the La Victoria property and will provide the exploration company with vital information about both areas of interest.

Object Recognition for Large-Scale and Weakly-Labelled Medical Image Data

The main objective of this research project is to investigate, develop and evaluate state-of-the-art image processing and machine learning algorithms, which are suitable for accurate modeling and recognition from large-scale medical image datasets that are weakly labeled. In particular, we will focus on the learning of recognition models in medical image computing applications that are of high interest/priority to Corstem, for instance, finding automatically the left and right ventricles in short- and long-axis cardiac magnetic resonance (MRI) sequences, which yields diagnosis measures that are of high interest to clinicians. Learning recognition models in medical image computing typically leads to difficult computational problems, where imaging data sets are weakly annotated.
This set of projects will leverage some limited and targeted interactions with medical experts, as needed, to set anatomical constraints and to drive advanced learning methods. TO BE CONT’D

Automatic Modulation Classification using Deep Learning for wireless security applications

Applications of wireless security approaches are increasing in number by the day. One such application is detection and interception of rogue aerial intruders. Drone technology is growing at a tremendous pace and is expected to be a $12 billion industry by 2021. Coupled with this growth comes the increasing threat of rogue intruders disrupting day-to-day activities and sensitive infrastructure. Towards this end, rogue drone detection has become an important industry by itself and focuses on securing the airspace around us. In this project, we aim to build a deep learning framework augmented with traditional signal processing techniques to model and classify unknown wireless signals. This model would be able to learn from the time domain information of the signal (amplitude, phase) and be robust in conditions with varying Signal-to-Noise Ratio (SNR). TO BE CONT’D

Statistical machine learning methods applied to ATB data for credit risk modelling

Machine learning (ML) is a method of training a computer to learn from data and predict future outcomes based on existing patterns in the data. This project aims to utilize various ML methods as new and potentially better analytics and predictive tools in the area of credit risk management for ATB. Given that data quality and flows change over time, a new framework built on Google Cloud Platform to update the machine learning models will also be developed. Additionally, considering the possibilities that the ML models may favour certain subgroup, e.g., defined by race and geography, new strategies to test and correct for model fairness will be established. In summary, we would improve default risk prediction accuracy with the help of leading edge ML methods and expand the field of credit risk management by providing model updates through Google Cloud Platform and establishing model fairness strategies.

Development of an RNAi Approach to Control Wireworms on PEI

Wireworm (click beetle) has becoming a devastating pest for the potato industry in Prince Edward Island (PEI) and many places in the country and around the world. It is a soil born insect that can penetrate a large number of potato tubers, degrading their quality for processing. The problem is more difficult to control after the ban of soil fumigation was in place in PEI. This proposed work is aimed at developing the RNAi technology by targeting the key genes for insects’ survival and development to control the insect population. The long term goal of the project is to deliver the RNAi molecules through the cover (bait) crops that are used to rotate in the potato fields. This specific proposal is the continuation of a two year project. To date, promising candidate genes have been cloned and DNA sequenced. TO BE CONT’D

Assessing the effectiveness of shorebird roost site conservation strategies through the employment of Open Standards for the Practice of Conservation

The purpose of this project is to conserve critical shorebird roost habitats at high tide in the Minas Basin of the Bay of Fundy in collaboration with recreational beach users, local businesses and tourism operators. The project seeks to develop long-term solutions for creating safe spaces for roosting shorebirds by identifying innovative strategies whose effectiveness can be measured using the Open Standards for the Practice of Conservation protocol. This work will allow BSC to develop an in-house capacity for the protocol as well as a proof of concept guide.

Exploration des technologies innovantes via le traitement du langage naturel et l’apprentissage machine

L’amélioration est processus (quelle que soit leur nature) est un souci constant des entreprises dans un domaine compétitif comme c’est le cas du domaine bancaire et financier. En 2018, cette amélioration est indissociable de l’intégration de l’analytique d’affaire. En exploitant à l’aide des méthodes appropriées les données de l’entreprise, des gains substantiels sont à escompter.
L’objectif général de cette grappe est d’exploiter les données disponibles dans les bases de données de la BNC afin de pallier aux lacunes dans les connaissances scientifiques et professionnelles sur le sujet. Selon les sous-projets, les bases de données RH, les bases de données clients ou les données opérationnelles de fonctionnement seront exploitées.

Developing an Intelligent Conversational Agent Architecture related to the Banking Domain

This research project aims at creating a robust, efficient and reliable conversational agent for the banking domain that will offer a high level of performance in both key areas of conversational agent architecture: Natural Language Understanding and Response Generation.
Natural language understanding approaches, retrieval-based models, as well as deep learning will be used to develop the architecture of the conversational agent in this specialized domain.

Quantifying fish-turbine interactions using VEMCO’s new high residency acoustic electronic tagging technology

It remains unclear if fishes that occupy Canada’s leading tidal energy test site in Minas Passage, Bay of Fundy, Nova Scotia, will be negatively affected by turbine installations. The objective of this project is to determine fishes’ interactions with operating turbines. Of approximately 70 species of fishes that interact with Minas Passage only three have abundance estimates (Gaspereau River Alewife; Shubenacadie River Striped Bass; Saint John River Atlantic Sturgeon) that are necessary to predict effects at the population level. Atlantic salmon are listed endangered by SARA so the loss of even one individual is significant. We will use new, innovative High Residency fish tracking technology from industry partner VEMCO to determine spatial and temporal overlap, and interactions of Alewife, Atlantic Salmon, Striped Bass and Atlantic Sturgeon, with operating turbines. This will provide information central to predicting negative effects of operating turbines, on fishes, at the population level.

Modeling and Dynamic Performance Assessment of a Battery Energy Storage Systems

Bulk storage of energy is a relatively new concept in many power systems. Among various energy storage media, batteries have shown great promise as a suitable option for use in power systems. Integrating a battery energy storage system in a power grid is not a trivial task and requires extensive studies to ensure that the system is able to respond satisfactorily to its surrounding’s variable conditions and deliver what is expected of it. Additionally, batteries experience marked changes in their characteristics as they age, which may adversely impact their performance and even cause undesirable outcomes such as instability. This project aims to study battery energy storage system dynamics and the impact of aging using advanced computer simulation and modeling techniques.

DC Interconnection Hubs

Conventional power systems are based upon ac voltages and currents. Connecting these systems is a simple matter and is done using transformers. Modern power systems wherein renewable energy sources are increasingly deployed often include dc voltages and currents. Connecting these systems is more challenging as conventional transformers will not be applicable. The proposed research is aimed at investigating and evaluating options for linking and interconnecting dc power systems. Power electronics is the enabling technology for achieving dc system interconnections. Various dc interconnection schemes will be researched and their merits and drawbacks will be thoroughly assessed using detailed computer modeling and simulation techniques as well as experimental work on a scaled down laboratory setup.

Improving Combustion Properties of Energy Crops and Agricultural Residues through the Removal of Undesirable Nutrients

CENNATEK Bioanalytical Services Inc. is leading a R&D project with the goal of

producing an improved biomass fuel for combustion through a pre-treatment process

involving the extraction of undesired nutrients. The presence of these nutrients often

contributes to adverse impacts on reactors and other equipment, with fouling, slagging,

and corrosion the most problematic of these impacts. CENNATEK has developed a

liquid extraction process that will remove these nutrients from the biomass, while at the

same time recovering and concentrating the nutrients to produce a liquid fertilizer to be

returned back to the soil. The proposed project would optimize the process on a labscale,

followed by modelling and designing of a pilot-scale process. After construction,

the pilot-plant will be optimized and data gathered to develop a full-scale commercial

plant that will be both technically and economically feasible. CENNATEK has an

aggressive time-to-market plan for the commercialization of this technology, and will

benefit through the potential IP and know-how generated from the project’s success.