Projets novateurs réalisés

Performance issues for embedded and random problems in adiabatic quantum optimization

Current and near-future D-Wave processors implement optimization using quantum effects on a fixed processor layout. This layout can be modeled as a graph with certain constraints, and the optimization is applied to a specific problem: fixed-topology Ising spin optimization. It is therefore important to study how best to use this hardware to solve combinatorial optimization problems in various forms, and to study performance versus conventional optimizers. The intern will study problems in this area, as well as the problem of predicting performance of proposed processor topologies. D-Wave expects to benefit from this partnership by gaining a clearer view of these issues, which are of fundamental importance to designing software applications for quantum annealers.

A Comprehensive Approach to Automated Essay Scoring through Weak Supervision, AutoML, and Interpretability

Scoring essays is important for learning and business. With AI and machine learning, we can save time and resources by automating the process. This project looks at automatic essay grading from three angles. First, we generate more data using automatic methods called data programming. This helps improve the machine learning models. Then, we refine the data using NLP techniques. Second, we will use Auto Machine learning techniques to easily determine the parameter for training. Third, we make the model more understandable. Deep learning models are often hard to understand, so we’ll use methods to explain the decisions made by the algorithm. We’ll focus on local and post-hoc explainability models. For data programming, we’ll use weak supervision. For interpretability, we’ll use methods like SHAP and LIME.

Methodology to Collect, Calibrate and Process Thermal RPAS Imagery into Radiometric Orthomosaics for Salmon Habitat Preservation

This project proposes to create a standardized methodology to collect, calibrate and process thermal drone imagery into ortho-rectified maps showing accurate temperature data for streams and rivers. These temperature maps are quickly becoming a useful tool with which to detect influxes of cold water that are critical for the preservation of salmon habitat. As our streams and rivers become increasingly hotter due to the effects of climate change, it imperative that new, efficient methods are developed to identified stream reaches that help to cool the stream and preserve salmon-rearing habitat.

PandoPartner Internship Project

This project will help PandoPartner train its clients on how to use our software to better their sponsorship marketing operations, as well as help PandoPartner promote its software. It will involve the creation of User Guide Content & client resources, Marketing materials, and more to help clients and prospects understand how they can leverage PandoPartner’s software.

Prey Availability and Predation as Limiting Factors of Burrowing Owl Population Growth in British Columbia

Alison’s research, conducted in partnership with the Burrowing Owl Conservation Society of BC, aims to better understand what is limiting the endangered burrowing owl’s population recovery in British Columbia. The Burrowing Owl Conservation Society of BC has implemented a species reintroduction program that includes breeding and releasing owls. They then track how many owls survive, reproduce, and return from migration. Alison’s research will focus on understanding whether release site prey availability and predation rate are potentially limiting burrowing owl reproductive success and survival during the breeding season. Analysis of data collected via wildlife cameras, insect sampling and direct observations will provide critical information on how prey availability and predation rates differ between release sites. This information will be useful in guiding where future releases occur, as well as making informed management decisions regarding the habitat currently used to maximize burrowing owl population growth.

Feasibility and Initial Efficacy of Remote Home-Based CardiacRehabilitation— A Pilot Study

Cardiac rehabilitation (CR) is proven beneficial for patients with cardiovascular diseases. Currently, most healthcare systems in China provides limited CR program toward patients after the cardiac intervention. This is particularly true among the rural areas. Home-based cardiac rehabilitation (HBCR) is used to alternative center (hospital) based in many developed countries. In this proposed research, a pilot study will be carried out to investigate the feasibility and effectiveness of using emergent mobile and information technology to deliver monitoring and feedback function in HBCR program. We aim at establishing the understanding the determinants of a successful home-based rehab program in terms of serviceability, reliability and adaptability of such mobile Health program in China. The findings from this pilot study are hoped to be compared with similar programs in Canada and other developed countries. The sensors used to monitor patients’ health indicators include Heart Rate (HR), Electrocardiography (ECG) and built-in geological sensors in smartphones. The rehabilitation programs will be customized according to the information collected from these sensors, and responses from the patient’s self-assessment. After the 3 months intervention, the participants’ health indicators and the program’s operational data are analyzed to assess the acceptability, feasibility and initial efficacy of the proposed HBCR program.

Stratégie de valorisation des techniques et technologies environnementales d’Écoboeuf

Peu de solutions sont accessibles aux agriculteurs pour adresser la problématique de l’adaptation et la lutte aux changements climatiques. Écoboeuf aspire à développer un portfolio de techniques et technologies propres applicables dans le contexte des régions périphériques boréales et à valoriser les réductions de gaz à effet de serres (GES) par la mise en marché de la production de ces fermes avec des marques alimentaires à valeur ajoutée environnementale (‘insetting’).
Le modèle de production d’Écoboeuf adresse simultanément plusieurs enjeux éthiques et environnementaux, ce qui résulte en une proposition de valeurs multifactorielle pour son produit. Cependant, les différents éléments de la proposition de valeur n’ont possiblement pas tous la même valeur aux yeux des consommateurs.
Ce projet vise à déterminer la meilleure approche pour capturer la marge de revenu supplémentaire atteignable par la commercialisation de produits alimentaires à valeur ajoutée environnementale. Une analyse de la valeur perçue des différents avantages du produit sera réalisée afin de cibler les composantes de la proposition de valeur ayant une forte propension à être payées des consommateurs. Les segments de clientèle sensibles aux éléments générant la plus forte propension de payer de la proposition seront ensuite identifiés et décrits. Une stratégie de distribution et de communication sera alors élaborée afin de choisir les canaux et messages les mieux adaptés pour ces segments de clientèle. Ce projet structura la stratégie de croissance d’Écoboeuf, participant ainsi à la transformation du secteur de l’élevage canadien et l’atteinte de ses cibles de réduction de gaz à effet de serre.

Empowering Literacy On-Demand with Interactive Multimedia

The project develops innovative delivery of beginning literacy lessons.
While learning to read and write stories of their own choosing, early learners have many urgent questions. Many of these questions have answers that can be easily explained using a multimedia demonstration. Children can access the answers independently, on demand, when the answers are in a digital repository with straightforward navigation.

In conjunction with outside educators and software developers, this project develops engaging multimedia explanations for Words, Sentences, and Texts, and tests the content with early learners, parents, and teachers.

Assessing the performance of an innovative IoT two-bin kanban “smart” shelve on a hospital’ logistics

While most of the research on RFID-IoT in healthcare sector focuses on mobile devices management and patient monitoring, this research contributes to our understanding of Internet of Things (IoT) technologies to improve (internal) hospital’ medical supply logistics processes.
Although various replenishment-distribution point of use solutions are available in the market to meet the expanding workload, little is known about their real impacts on reducing costs, improving availability of medical supplies and driving better clinical outcomes. Additionally, with various competing solutions and supporting technologies, there is no single dominant model nor “best” replenishment distribution method. Managers are then wondering what is the “most appropriate” option for their use cases. The comparative analysis is not simple, since most of solutions are sold as proprietary solutions (lock-in licenses with vendors) with the underlying technological design, often presented at a high level by the vendors.
Following a Design Science approach, the purpose of this project is to develop an IoT two-bin kanban “smart” shelve replenishment system of medical supplies that can be used in a hospital nursing unit. This smart shelf combines various technologies including passive ultra-high frequency (UHF) radio-frequency identification (RFID) technology and indicative led.
The originality of this project lies in combining RFID-IoT technologies and hybrid simulation modeling that uses data from a real project to explore the impacts of the technological solution on critical logistical business processes and its impacts on patient care as well.

Création d’un outil de modélisation géospatiale d’un réseau cellulaire

Networks, Economics & Strategy (NE&S) est une entreprise de conseil spécialisée dans les stratégies pour les secteurs des télécommunications et de la technologie. Ils possèdent une vaste expertise, allant de la planification des réseaux à l’analyse économique. Avec les changements rapides dans les télécommunications, une question clé est de savoir comment gérer automatiquement l’offre et la demande pour chaque réseau cellulaire. La solution proposée vise à développer un outil de gestion et de visualisation des réseaux communication, aidant ainsi à améliorer l’analyse et la modélisation géospatiale. Cet outil facilitera la prise de décision pour les décideurs du secteur.

Testing a new parameterization for barotropic mesoscale ocean eddies in an Arctic basin setup

Accurate modelling of the climate is of paramount importance in developing strategies to mitigate the effects of climate change. The ocean is a crucial component of the climate and simulating accurate ocean conditions in climate models is therefore extremely important. However, the energy required to reach a high degree of accuracy in modelling the ocean is large. Smarter ways of thinking are required to reduce the energetic cost of modelling the ocean while still retaining a high degree of accuracy. This project will test a new method for modelling the Arctic Ocean which has a low energy cost. We will compare the results of a highly-accurate, high energy cost model, with that of the new method to determine if this method can produce accurate results at a fraction of the energetic cost. This new method may be used in future climate models depending on the outcome of this project.

The genetics of species formation in Amazonian birds: the case ofthe Silvered Antbird

This project aims to study how species form in the Amazon using the Silvered Antbird as a model. This species has diverged into two morphologically divergent forms. Where these forms come into geographic contact, populations appear morphologically and genetically intermediate suggesting they form a hybrid zone. To investigate the extent of genetic mixing between these forms I will use a genome-wide sample of tens of thousands of genetic markers sampled between eastern and western forms where they come into geographic contact. This data will be used to compare patterns of genetic merging, and to search for genes that may be important in promoting species formation.