Innovative Projects Realized

Optimisation de la géométrie de convertisseurs de puissance laser interconnectés en série

L’équipe de Gwenaëlle Hamon développe des procédés de micro-fabrication pour les dispositifs III-V, incluant des convertisseurs de puissance laser qui sont des dispositifs composés d’hétérostructures à base de GaAs (structure VEHSA (1) (2)). Ces hétérostructures convertissent la lumière d’un laser incident en électricité. La structure VEHSA (Vertical Epitaxial Heterostructure Architecture) consiste en un empilement de jonctions PN de GaAs connectés en série, permettant d’augmenter la tension des dispositifs. Le projet dans lequel s’inscrit ce stage consiste à interconnecter horizontalement ces structures en série, afin d’augmenter davantage la tension de sortie de ces dispositifs. Un des défis de cette interconnexion concerne l’harmonisation des courants. Chaque cellule doit fournir un courant identique. Ainsi, la disposition des cellules et leurs dimensions doivent être adaptées au profil de la lumière incidente du laser. Ce projet vise donc à poursuivre le développement de cette technologie en optimisant la géométrie des convertisseurs de puissance laser interconnectés en série afin de permettre une génération uniforme du courant électrique.

Data-driven simulation of DEVS-driven Digital Twins for smart manufacturing and Industry 4.0

Traditional modelling and simulation involves a human expert to manually design and create simulation models. However, these models quickly become obsolete for systems that continually change, such as smart manufacturing systems. This creates a need to create new simulation models which is difficult and expensive. Data-driven simulation extracts simulation models from data without explicit modelling from an expert. Digital Twins, virtual representations of physical objects, are considered one of the pillars of Industry 4.0. The DEVS formalism is an ideal implementation for the underlying Digital Twin simulation models due to their capability to model heterogeneous and complex systems. However, further research is needed in the development of algorithms to automatically extract the simulation models implemented in Digital Twins. This project focusses on extracting DEVS models using data-driven simulation with motivation in DEVS-driven Digital Twin applications for smart manufacturing and Industry 4.0.

Integrating THz Spectroscopy with Cryogenic Systems for Quantum Material Characterization

This Mitacs Globalink research project is aimed at advancing cryogenic quantum technologies through the development of a Terahertz (THz) spectroscopy test setup compatible with Adiabatic Demagnetization Refrigerators (ADR). The project will leverage the novel capabilities of THz spectroscopy to probe molecular dopants trapped in cryocrystals, enhancing the understanding of quantum materials and their low-energy excitations.

The intern, a postdoctoral researcher specializing in hybrid quantum systems, will join this cutting-edge research effort. Over a 12-week internship, the focus will be on integrating THz spectroscopy with existing ADR systems, optimizing this setup to function at the extremely low temperatures crucial for quantum research.

The research will occur in sophisticated lab environments, providing the intern with access to state-of-the-art THz and cryogenic equipment. This collaboration will not only bolster the intern’s expertise in quantum technologies but also foster international partnerships and knowledge exchange, strengthening Canada’s position at the forefront of quantum research and technology.

Globalink Research Award Application – Internship in Switzerland (May 1 2025 start) – McKinley Van Klei

Standard medical imaging techniques used to diagnose and inform treatment of spinal conditions include X-Ray, computed tomography (CT), and magnetic resonance imaging (MRI). These images neglect to capture the in vivo dynamic behavior of the spine during activities of daily living, since patients remain as still as possible to capture clear image. In 2023, 20% of the adult population was reported to be diagnosed with spondylolisthesis, a condition that can cause debilitating low back and leg pain. The diagnostic process and treatment of lumbar degenerative spondylolisthesis (LDS) is controversial and lacking in evidence, however common treatment is spinal decompression or fusion surgery. The purpose of this study is to provide clinically relevant insights into the biomechanical markers of LDS using DBRI, with the goal of improving diagnostics and informing patient-specific treatment plans. Patients diagnosed with LDS will be examined pre- and 6-months-post- op using DBRI, EMG, force measurements, CT, and MRI (target n = 100). DBRI will be conducted while patients perform a dynamic flexion extension motion using the established facilities in the Digital Imaging Centre at sitem-insel. This study will contribute to the “translational musculoskeletal biodynamics” initiative by discovering biomechanical markers and biomechanical profiles unique to LDS using DBRI.

An efficient implementation of a computational camera for smartphones

A camera is a device that captures light from scenes. Over the last century, the evolution of cameras has been truly remarkable. Through this evolution, the underlying camera has been improved by using a better optical lens. However, the new improved optical lenses, have been remained fixed in terms of size and weight which makes it hard to use in portable devices. In contrast to optical trend, according to Moore’s law, the number of transistor in the chip doubles approximately every two years. This leads to a huge improvement in computational devices. The slim smartphones with powerful processors are an example of this improvement. This motivates using more computation rather than better optical equipment in the structure of cameras. This results in a new generation of cameras called computational cameras. This project aims to implement a new computational camera for a smartphone. The new implementation has mainly designed to reduce the blurring effect caused by hand shake or the lens problems. Moreover, the implementation of this project is optimized for a smartphone such that it performs in real time with a low power

Enhancing snoRNA family annotations in the Rfam Database

This project aims to enhance the annotation of small nucleolar RNAs (snoRNAs) in the Rfam database, a key resource for classifying non-coding RNAs (ncRNAs). Rfam groups ncRNAs into families based on evolutionary relationships, using initial manual curation followed by a computational pipeline developed by the Rfam team. SnoRNAs are ncRNAs that guide RNA modification and are linked to diseases like cancer when their functions are disrupted. In vertebrates, many snoRNAs exist in multiple copies, potentially acquiring novel targets or functions, highlighting the need for accurate family classification to understand their regulatory mechanisms. Advancements in experimental and computational methods continue to reveal novel snoRNAs, further emphasizing the importance of updating snoRNA family annotations. This project seeks to expand and refine the Rfam database by identifying and organizing previously unannotated snoRNAs. These updates will improve Rfam’s accuracy, better reflecting snoRNA-specific characteristics and providing the ncRNA community with more reliable datasets for studying snoRNA functions and their disease associations. The improved Rfam snoRNA families will specifically benefit the Scott lab (home institution) by providing new Rfam IDs for unannotated snoRNAs, enriching snoDB – a specialized database of human snoRNAs – and supporting further research into snoRNA diversity and function.

Quantum algorithms for non-adiabatic dynamics

Xanadu’s mission is to make quantum computing useful, through development of quantum hardware, software, and algorithms. One important direction in achieving this goal is identifying problems that can be solved on quantum hardware that are not tractable on classical computers, and then building quantum algorithms for those problems. We expect that eventually we will have access to a device that has a small number of error-corrected logical qubits, with the ability to implement complicated circuits on those qubits. Thus, to get practical use out of all the hardware improvements, it is important to turn our attention to quantum algorithms specifically designed for these early fault-tolerant quantum computers.
One promising direction Xanadu has identified is the simulation of non-adiabatic dynamics. Advancing and extending our recently developed framework for these simulations to cover a broader range of applications is essential to our mission of making quantum computing practically useful. By refining these methods, we aim to unlock new capabilities that leverage early fault-tolerant quantum hardware for problems beyond the reach of classical computation.

Caregiving skills for the engineering profession

This study explores the relationship between caregiving skills and characteristics and the engineering profession. The goal is to understand how caregiving experience contributes to the development of skills that are beneficial in engineering organizations. Interviews and surveys will be conducted to: (1) identify caregiving skills relevant to engineering, (2) explore how these skills are applied in the workplace, and (3) examine whether caregiving experience enhances professional development. The study aims to explore whether caregiving can offer transferable skills that benefit engineering organizations. Additionally, results are intended to support women in engineering, a field where they are underrepresented and often face more intense caregiving responsibilities. This research seeks to contribute to discussions on caregiver support and demonstrate the value of caregiving while working as an engineer. This study also intends to support Engineers Yukon’s continuing professional development program, where caregiving is identified as a professional development activity.

Technical Business interns working within cross-functional teams to commercialize AI-powered solutions in the Public Services Sector (1)

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

Blockchain Smart Contract Vulnerability Detection Using Quantum Convolution Neural Network

The proposed project aims to develop a Quantum Convolutional Neural Network (QCNN)-based approach to detect vulnerabilities in smart contracts, which are critical components of blockchain technology. By leveraging quantum machine learning techniques, the project seeks to enhance the accuracy and efficiency of identifying security threats in smart contracts, such as reentrancy attacks and integer overflows. This research will involve collecting and preprocessing a dataset of smart contracts, designing and implementing a QCNN model using frameworks like TensorFlow Quantum and Qiskit, and benchmarking its performance against traditional deep learning methods. The expected benefit to the participating institutions includes advancing their expertise in quantum computing and blockchain security, fostering collaboration between researchers, and contributing to the development of more secure and reliable blockchain ecosystems. This project will also provide valuable training opportunities for interns, equipping them with cutting-edge skills in quantum machine learning and cybersecurity.

Caractérisation des fibres de bois recyclées

Il est possible de recycler les panneaux de fibres à densité moyenne (MDF) par un procédé hydrothermique, décomposant la résine thermodurcissable urée-formaldéhyde liant les fibres de bois le constituant. Bien que ce procédé soit de plus en plus connu, nous disposons de peu d’informations sur les propriétés chimiques des fibres récupérées, ce qui a pourtant un impact crucial sur les différentes applications possibles de ces fibres récupérées. Ce projet a pour but d’étudier les propriétés chimiques, principalement la proportion des constituants (composants lignocellulosiques et résine résiduelle) des fibres de bois récupérés par procédé hydrothermique et d’évaluer comment l’utilisation d’acide faible et la température influenceront la proportion des différents constituants des fibres. À terme, ce projet propose de répondre à certaines problématiques que rencontre le Canada, comme le recyclage de panneaux de bois MDF et la valorisation d’acides faibles industriels.

Exploring Quantum Computing for Public Transit Origin-Destination Matrix Estimation

This research explores the integration of quantum computing with statistical methods to enhance public transit origin–destination (OD) matrix estimation. Traditional OD estimation relies on automated fare collection (AFC) and automatic passenger counting (APC) data, which often present challenges due to incomplete coverage. Scaling techniques like iterative proportional fitting (IPF) help address these gaps, but their accuracy declines at low AFC penetration rates.
This study proposes leveraging a hierarchical Bayesian framework alongside quantum algorithms, Quadratic Unconstrained Binary Optimization (QUBO) for combinatorial optimization and the Harrow-Hassidim-Lloyd (HHL) algorithm for solving linear systems, to improve OD matrix accuracy. These methods will incorporate APC-derived alighting probabilities and AFC trip-chaining data. The approach will be validated on the Sioux Falls and Calgary transit networks, assessing its robustness under varying data conditions.
This project, led by Dr. Saidi and Dr. Nassir, aligns with Canada’s strategic priorities in quantum technology and transit innovation. By enhancing OD estimation, it aims to improve transit planning, optimize operations, and contribute to more efficient urban mobility. As the intern, I will gain hands-on experience in quantum computing and transportation analytics, fostering collaboration between Canadian and Australian research institutions.