Projets novateurs réalisés

Efficient Adjoint Sensitivity Analysis of Emerging Nanophotonic Devices

The design optimization process of photonic and optoelectronic devices can be time consuming. The modeling of these electrically-long devices is time-intensive. Accurate time-domain and frequency-domain simulation tools such as the finite difference time domain (FDTD) method and the finite element method (FEM) are usually utilized to model these devices. These methods require fine meshing to achieve accurate results. The fine meshing results in slow simulation time.
Designing photonic and optoelectronic devices requires large number of accurate simulations. The robust gradient-based optimizers require repeated estimation of response sensitivities. The slow models of electrically-long devices render the optimization phase formidable. Several approaches have been proposed to accelerate the design of these devices. These approaches including surrogate modeling approaches such Space Mapping. They also include adjoint sensitivity approaches. These methods evaluate the sensitivities of the objective function in an efficient way thus dramatically reducing the optimization time.
This project will provide selected interns with the opportunity to develop fundamental analysis and numerical software development skills for a broad range of industrially relevant physical simulation challenges. Further, the project will provide R&D services to the industrial partner, leading to advanced research, potential new products and opportunities for Canada.

Predictive Maintenance Platform Development for UAVs

This project aims to develop a solution for the predictive maintenance of UAVs which will mitigate the burden and risks associated with the traditional reactive and preventive maintenance. By identifying the critical data related to the failures and monitoring/modeling the behaviors of the components based on data/signals during and after the flight, this research will develop the capability to identify the affected component in advance in a predictive manner. It will not only increase the safety of the UAV operation but also reduce the risks for the organization, and at the same time minimize the costs for the in-house maintenance and system check. This project aims at the analysis and development of AI-based predictive models thanks to the data collected during the use of RPAS, in order to generate a health score of its components and hardware.

Improvement of GCxGC-MS compound identification and development of multi-omics integration techniques using a clinical stool sample set

Multiomics studies of the human microbiome have an enormous potential for understanding the mechanism of biological processes involving the microbial community and the metabolites produced by gut microbiota. We will apply a multiomics strategy to clinical fecal samples using a novel collection device to stabilize fecal metabolites at room temperature, thereby making the standard practise of freezing at – 80°C unnecessary. Room temperature storage allows at-home collection by donors/participants, enabling large-scale studies to take place, which are needed for research discoveries to further our understanding of the human gut microbiome and metabolic pathways involved in health and disease. This study will demonstrate feasibility and utility of DNA Genotek’s collection tool for use in multiomics studies. In addition, the multiomics data analysis will provide insights into the complex biological interactions and metabolic processes taking place in clinical samples.

Multiculturalism and Canada’s Stage: Assessing the diversity of the National Arts Centre’s English Theatre Program, 1969-1994

This project will analyse Canada’s official multiculturalism policy and its impact on the National Arts Centre’s
English language theatre program. The National Arts Centre Archival holdings will be utilized for research to
assess the diversity of persons from ethnic minorities and Indigenous groups as actors, directors, and authors of
the plays that have been performed through the twenty-five years following the opening of the National Arts
Centre. The findings of this research, as well as a detailed history of the National Arts Centre and Canada’s official
multiculturalism policy will be published within an online digital exhibit on the National Arts Centre’s website
and will be made available to the public.

Modeling and real-time optimization for sustainable and robust smart freight transportation platform

Urban freight transportation plays an essential role in the future of smart and sustainable cities. The emerging of e-commerce, new business models, and evolving technologies will affect logistics efficiency, policymaking, and planning in the freight transportation context. Smart freight platforms will be the next generation of shared mobility in global supply chain and logistics. It contributes to the sustainability of cities and all stakeholders involved in these platforms. This research project aims to develop the optimization models and learning based solutions for matching, pricing, repositioning and routing in this platform under uncertainties.

The Impact of Virtual Background Characteristics on Asynchronous Video Ratings for Health Professions Education Admissions

The COVID-19 pandemic has presented numerous challenges for Health Professions Education programs, especially those that needed to switch from in-person to online interviewing as a means of evaluating applicants. Particularly, despite the rising use of online formats such as asynchronous video interviews (AVIs) in selection decisions, there is little evidence on the fairness and extent of bias associated with these tools in admissions contexts. The proposed project will examine how various AVI characteristics (e.g., applicants’ choice of interview location, attire) impact their evaluations. This objective will be accomplished via the intern working with Altus Assessments, who developed an AVI for admissions to programs. This research will provide not only scientific evidence on the potential biases underlying the rating of AVI responses, but also the specific interview tool (Snapshot) that Altus Assessments developed. Ultimately, this research will help guide best practices in offering a fairer test which Altus Assessments can implement.

Improved Augmented Reality (AR) Alignment for Industrial Steel Fabrication Process

Fabstation-Steel is the flagship product of Eterio Realities Inc. It is designed to improve Structural Steel Fabricators’ production quality and velocity by using AR (augmented reality) virtual assembly model overlay technology. Since the accuracy of the virtual model overlay is critical to the success of the product, this research project aims to improve the virtual model overlay accuracy performance of the Fabstation-Steel system by applying advanced computer vision tracking technology. Any improvement to be achieved in the Fabstation-Steel overall performance will streamline and enhance the productivity of the steel fabrication process, retain the current Fabstation-Steel users, and attract new customers. Ultimately, this project seeks to improve working conditions for structural steel workers by improving data delivery and cognition which will reduce production waste and associated emissions. This aligns with environmental sustainability practices being promoted across the country.

Development of Innovative Connections for Shelf Angels Used in Wood-Framed Building

Most residential houses in Canada are built with wood frame. Brick veneer is installed on the outer surface of the wood building to provide an elegant brick-finish look. Steel shelf angles are mounted on the word frame of a building to support the brick veneer’s load. Installation of shelf angles requires installation of through bolts which need to pierce through the wood frame into the building envelope creating thermal bridges. These thermal bridges allow internal heat (in winter months) and cool air (in summer months) of the homes to escape to the outside atmosphere and thus, a large portion of the energy used for heating and cooling is lost. Hence, this research will develop thermally efficient connections for mounting shelf angles. The goal is to reduce the loss of energy in our homes and help Canada in achieving the target of net-zero energy buildings by 2050.

The Molecular-Rich Outflow of Giant Stars

When stars like our Sun reach the end of their lives, they become giants and eject a lot of their initial mass through a slow, dense outflow. These winds are driven by microscopic solid particles (dust grains) that condense in the cool stellar atmospheres. The dust is the product of a large number of chemical reactions that take place in the extended atmosphere and outflow, and which also lead to the formation of several molecules. By studying the abundances and distribution of these molecules, a better understanding of the chemistry and physical properties of the ejected gas can be achieved. Ground-based observatories such as ALMA now allow us to detect a large number of molecular lines and map the emission regions in an unprecedented way, providing crucial new constraints to the study of the chemistry in the outflows of evolved stars. This CASSUM project concerns the study of observations of a sample of several tens of evolved stars observed with the Atacama Compact Array.

Develop a Rule-based Entity Recognition Model for Unstructured Data in Artificio Platform

Preparing and analyzing unstructured data from various documents (e.g. leasing agreements, engineering
drawings, etc.) is time consuming and difficult task for any organization or businesses. Artificio platform developed/created by ERP Solutions2go (dba BizTech2Go) will help the businesses or enterprises in automating this tedious process, the manual data entry process , and seamlessly integrate the same into their ERP Systems. Artificio platform will leverage various AI/ML methods and techniques for text detection, text recognition and entity extractions. Artificio will provide very simple UI backed by the pretrained models for different applications in different industry sectors so that users can adapt the AI/ML models with zero development or zero code. The deliverable will be a Rule-based Entity Recognition Model for Unstructured Data and Present LayoutLMv2 by pre-training text, layout and image in a multi-modal framework, where new model architectures and pre-training tasks are leveraged.

Engineering long range interactions for superconducting quantum computers

This project aims to provide a solution to a critical problem in existing superconducting quantum processors by developing methods to create long-range interactions. Superconducting qubits are one of the most promising platforms for building quantum computers. Currently, quantum processors based on this technology offer the largest quantum memories, but do not benchmark well in terms of the size of the quantum program that can be executed. One of the main drawbacks of the existing platforms is that the qubits only interact with their nearest neighbors on the chip. By finding methods for creating effective long-range interactions between qubits, the size of the quantum programs that can run can be larger. Entangled Networks works on providing hardware and software solutions to enhance the performance of quantum computers and will incorporate the results as part of their optimization package.

A Techno-economic assessment of renewable natural gas (RNG) as a potential low carbon intensity fuel alternative eligible under Canada’s Clean Fuel Standard

The aim of this project is to determine whether renewable natural gas (RNG) is a feasible low carbon intensity (CI) fuel alternative which can be produced using existing production technologies (anaerobic digestion) and farm waste (corn silage and/or wheat, barley, oat straw) at an economically feasible cost within Canada. Driven by the introduction of Canada’s Clean Fuel Standard (CFS), this work will also aim to determine the potential value which may be generated as a result of the establishment of a national low carbon fuel credit market, and the role this national market plays in determining the profitability of RNG production. This investment opportunity will be compared against alternative credit generating opportunities defined under the CFS to determine the relative costs of credit generation under unique conditions and the market sizes which can be developed within Canada for RNG and the associated credits. Dr. Jennifer Winter will act as the Academic Supervisor for this project and TransAlta’s Braydon Boulanger & Lukas Hansen will serve as Industry Partners.

The partnering organization will benefit from this work as it will offer unique insights into the potential profitability of an investment into RNG and the CFS. Specifically, it will outline the