Projets novateurs réalisés

Development of context-aware data analysis framework

In a complex environment like mineral extraction, it is difficult to back the cause-effect relationship between a specific process variable and the overall performance (output, cost, GHG emission) with scientific principles. Consequently, process data is used to deduce an empirical correlation between the process variables and their estimated effects. In the present project, process data from a mineral extraction plant will be used to analyse the cause-effect relationship of various plant variables on critical performance metrics like energy consumption, cost, GHG emission etc.. Furthermore, the analysis will factor in the contextual variables like environment, labour, market demands, etc. to make the framework adapt to dynamic changes. The project will advance NTWIST’s ML and AI based data analytics framework to provide superior solutions to their clients. The project will simultaneously help the Canadian mineral industry to enhance their efficiency, and the Canada goal of reducing GHG emissions.

Crabs and Community: A Co-Created Assessment of Early Dungeness Crab (Metacarcinus magister) Life Stages in Tsawout First Nation’s Waters

Dungeness crab occupy three distinct habitats over their complex life cycle as they progress from larval to juvenile to adult stages. To understand Dungeness abundance then, attention needs to be paid to each life stage and its
unique interactions with its environment. In the Salish Sea, studies addressing early Dungeness crab life stages, such as the final larval (i.e., megalopae) and juvenile stages are lacking. In this project, we attempt to address
these knowledge gaps by monitoring Dungeness size and abundance at these two early life stages using light traps and intertidal surveys. This project will be co-directed and co-conducted by the Hakai Institute, the University
of British Columbia’s Centre for Indigenous Fisheries, and Tsawout First Nation, focusing on study sites identified by Tsawout community members. By working together to assess early Dungeness life stages and the potential
environmental regulators of Dungeness size and abundance, this project will complement broad-scale work being conducted by Hakai, and will meaningfully support Tsawout in developing a Nation-specific Dungeness crab
stewardship plan.

Question Answering over Long Clinical Documents

In traditional claim processing system, adjusters or claim handling specialists need to read massively number of medical reports like diagnosis report, medical assessment, prescription daily for different types of decision making. Such a manual review process is time-consuming and tiring, which causes slowness in serving claimants and consequently unpleasant customer experience. In the digitalization era, we aim to digitize the manual process of reviewing documents with NLP techniques. In this project, we aim at building an intelligent NLP system that can help the adjusters navigate and extract key information from the medical documents to accelerate the process. With the new system, the practitioners can ask questions and chat with the system to obtain desired information. This will enable adjusters to make proper decisions faster leading to better claimants’ experiences.

Improving Anti Money Laundering through Active learning Methods

Fighting against financial crime is a battle that humanity is losing as a whole. Money Laundering is the critical component of any criminal activity that needs to be profitable. Therefore, fighting money laundering is fighting against all crimes including terrorism, human trafficking, drugs and wildlife trafficking. The United Nations (UN) estimates 2 trillion USD is laundered globally, and only less than 0.2% of that can be recovered by the authorities. The main responsibility lies in the financial institutions, where the majority of the laundering activity occurs. However, these institutions use rule-based systems that generate huge amount false-positive alerts and fail to catch criminals. Machine learning is a promising tool that has shown good results for complex tasks such as this one. In this project we propose to tackle the problem of money laundering detection through active learning strategies, where a human expert provides labels for cases carefully selected by the machine learning algorithm. We will identify the most well suited active learning strategies for the problem using multiple types of data including individual customer data and sub-graphs. A new special-purpose active learning solution will be developed to achieve better performance for this problem

Reaction-diffusion equations and hydrodynamic limit of interacting particle systems

Consider a physical system comprising a substantial quantity of particles that dynamically interact with each other over time. Since the number of particles is large, an explicit description of the microscopic behavior of the system becomes effectively impossible. An approach to understanding how these systems behave is to look at certain observables on a macroscopic scale, such as pressure, temperature, etc., and then to derive partial differential equations known as hydrodynamic equations which describe the evolution of the particles. In this project, we study the theoretical mechanism by which the microscopic dynamics give rise to the macroscopic ones. In particular, we are interested in systems whose hydrodynamic limit corresponds with an important class of equations known as reaction-diffusion equations. We expect to identify particle systems with the aforementioned property, and then rigorously establish the convergence to its corresponding hydrodynamic limit.

Cooperative Inversion of UAS-based Magnetic and Radiometric Observations in Mineral Exploration

Mineral exploration is one of the biggest drivers of the Canadian economy, and the critical minerals discovered in Canada
have global relevance for a green energy market. However, today’s mineral exploration projects require higher spatial
resolution and better subsurface models to become viable. Uncrewed aerial vehicles (UAV) present new exploration platforms
that carry geophysical sensors and acquire radiometric and magnetic data closer to the target than any other airborne platform.
This enables new modelling opportunities towards improved subsurface models. This study will combine radiometric data and
magnetic data in a cooperative inversion model to quantify how the acquisition of both data types improves mineral exploration
success. This project will result in innovative approaches which will keep Canada at the forefront of developing geophysical
exploration methods.

Evaluating mammalian diversity across a gradient of disturbance in Tsay Keh Dene Territory using eDNA

This project aims to identify the effects of human activities, particularly logging and mineral exploration, on mammals that live in the Tsay Keh Dene (TKD) Nation Territory. We plan to use a method called environmental
DNA (eDNA), which involves detecting genetic material from animals in the environment, to understand how much logging and at what distance logging activities influence the diversity of mammal species downstream in river
networks. The research will provide information about the different types of mammals in the TKD Nation area and will be useful for future land management and conservation planning. It will help the TKD Nation make decisions
about where and how much resource extraction should happen on their land. Moreover, this project will also help evaluate how effective eDNA can be as a tool to monitor the impact of disturbances on native fauna.

Vibration serviceability performance of mass timber slabs with glulam beam supports and concrete toppings

This research addresses the vibration performance of mass timber floors, which is crucial for their structural design and maximum allowable span. As mass timber construction is relatively new, there is a lack of specific design
guidance, especially for floors with beam supports and concrete toppings. Our goal is to assess the vibration serviceability of a large-scale mass timber floor with glulam beam supports using vibration testing and subjective
evaluation. The outcomes will provide valuable data for the industry partner to support the design of vibrationresistant mass timber floors. Additionally, our research will contribute to the wood construction industry, potentially
influencing future revisions of the Canadian engineering wood design standard CSA O86. By enhancing our understanding and design capabilities, we aim to improve the overall performance and guidelines for mass timber
floors.

The role of automobility engagement in consumer response to new mobility innovations

The transport sector is the second highest source of greenhouse gas (GHG) emissions in Canada. The prevalence of car use in passenger transport, also referred to as “automobility”, is a key driver of GHG emissions and other societal issues. Transport innovations can reduce emissions – for example, plug-in electric vehicles (PEVs) are an essential component in achieving climate goals. However, automobility patterns may still represent a challenge for transport sustainability. The Sustainable Transportation Action Research Team (START) and Navius Research Inc. are leaders in analyzing impacts from transport innovations in Canada. This project analyzes data from a large-scale survey of Canadian citizens to investigate consumer preferences for PEVs and other innovations, and their relationship to mobility by private cars. The project aims to improve knowledge of sustainable transportation pathways from a consumer perspective, which may help to inform transport and climate policies.

Microbial Dyes

Lite-1 utilizes innovative technologies like synthetic biology and biofabrication in pursuit of a novel solution. Many bacteria naturally produce colourful, non-toxic pigments (Ahmad, et. Al, 2013). Through the lens of design, we identify and study organisms that produce the market’s desired colours. Lite-1 dyes reduce water pollution, are safe for workers as well as the ecosystem, and eliminate our dependence on non-renewable resources. The product offers an overhaul to the dye industry to shift from one of the preeminent polluters to a regenerative and more humane industry.

This project is focused on leveraging the expertise and highly specialized equipment at the Lee Lab to improve the productivity of the Lite-1 organisms to achieve higher yields, streamline production and downstream processing protocols and decrease culture times through genetics.

RevMate: Towards an optimized review process in AAA game development

In software development, code reviewing is an essential code validation step during integration [1, 2]. When developers submit their code for review, human reviewers need to validate the changes made to source code
and other artifacts (e.g., data files, artifacts created by artists). This review process can go through several iterations in which the reviewers comment on the submission and the submission’s developer revise the
submission by submitting a new version (“revision”) addressing the suggested changes. Eventually, the (revised) submission either is accepted and forwarded to integration in the main development branch, or
rejected.
In this research project, we study different automation approaches to improve the velocity of the code review process at Ubisoft, our industrial partner. More specifically, we intend to build techniques for code chunk
ordering, hot-spot detection and review comment prediction. We will also establish a pre-review protocol where the developer validates and/or modifies our approaches’ results, as to have actual developers’ input and insights
in the review process. We aim to conduct this study on several AAA game productions, both ongoing and past games, to empirically assess the generalization of our results.

A High-Performance, Rapid-Cure Epoxy Resin Formulation for High Voltage Coils Using Radical-Induced Cationic Frontal Polymerisation (RICFP)

This project explores new polymer cure methods using a combination of light and thermally sensitive catalysts for epoxy thermoset resins. These resins can be used for a number of applications, including as insulators in high voltage coils. The postdoc and MSc interns will identify simpler epoxy formulations that utilize specially tuned photoinitiators and thermal catalysts. This will be accomplished by a design of experiments to identify potential formulation candidates followed by in-house testing at TMU and at the Trench site. The benefits to the partner is a simplified, more stable formulation with a much smaller carbon footprint and an increased throughput of epoxy coated electric coils during manufacture.