Innovative Projects Realized

Integrated modelling of tropical dry forest using climate, land cover change, and ecosystem services models

This project will combine specialized land cover, climate, and ecosystem services models into a single platform via the Dinamica EGO platform, linking them together using feedback systems. Using a deforestation model in Costa Rica as a baseline, with a climate change model, CORDEX, and a water services model, SWAT, we can produce a more complete picture of the future change in Costa Rica. This will provide me the opportunity to quantify how water resources will change under different land uses and climate alterations. Costa Rica is home to sensitive ecosystems, which have a high biodiversity and produce many services for nearby populations. Additionally, the economic activities in Costa Rica have historically centered on agriculture and ranching, and these contributions continue to comprise a large portion of the GDP. This focus on land based resources makes it imperative to understand how the future of the land will change, which areas will experience increasing demands on the water supply, and which natural areas will be resilient under climate and land use changes.

Les stratégies de pensée critique?des étudiant·e·s sur le Web social

Des fausses nouvelles aux the?ories du complot, la question de l’e?valuation de l’information est au cœur de l’actualite? sociale, e?conomique ou environnementale. En effet, pour certains, Internet et les re?seaux sociaux contribueraient a? la diffusion de rumeurs vis-a?-vis desquelles les jeunes et, par conse?quent, les e?tudiant·e·s ne seraient pas suffisamment outille?s. Pourtant, plusieurs de ces re?seaux, tels que Facebook, sont de plus en plus mobilise?s dans le cadre des e?tudes, y compris a? l’universite?. Concre?tement, ce projet a donc pour ambition de brosser le portrait des strate?gies qu’utilisent les jeunes lorsqu’ils e?valuent de l’information sur les re?seaux sociaux. De fac?on plus spe?cifique, nous nous concentrerons sur les indicateurs de « pense?e critique » que l’on peut identifier chez les e?tudiant·e·s qui entrent dans l’enseignement supe?rieur. Cette recherche s’inscrit dans un objectif de comparaison internationale (Canada, Belgique et France). Cela permettra d’entrevoir, e?ventuellement, des diffe?rences entre les populations cibles. Surtout, ce projet contribuera a? avoir un meilleur portrait de la re?alite? de l’e?valuation de l’information sur le Web et, finalement, a? mieux outiller les e?tudiant·e·s en conse?quence.

Hydroponic configurations and biological nutrient solutions for shipping container plant factories

Plant factories are closed growth environment, insulated, fully automated and artificially illuminated that can increase food security in regions such as Northern Canada. However. the technologies used in large scale plant factories are not suited for smaller scale plant factories, housed in shipping container. La Boîte Maraîchère design growth chambers in shipping container, but want to develop better hydroponic practices to get closer to the yields obtained in large plant factories. Hence, research is needed to increase the knowledge on small scale plant factories hydroponic configurations. This research concerns the comparisons of innovative multistage hydroponic configurations based on nutrient film technique, floating beds and aeroponics. The main innovation is to include the aspect of crop movement in these hydroponic system, to potentially automate the crop production.

Adaptive Radiation Therapy for Cervix Cancer with Stochastic Model for Tumour Evolution

The goal of this project is to develop new models to optimize the treatment of cervix cancer

using radiation therapy. We will develop two models to achieve this goal. The first model will

model the uncertain nature of the tumour evolution as a function of time, which will help us

predict the position, shape, and size of the tumour. This model will be validated using data

from our collaborators at Princess Margaret Hospital. The tumour evolution model will then

be used to support the development of a new adaptive optimization model to determine when

and how to re-optimize multi-week treatments. Treatments are delivered over multiple weeks

and the tumour’s geometry may change significantly over this time period. By incorporating a

model that describes the tumour changes into the optimization process, we will improve

treatments for patients, while reducing the use of expensive imaging resources at the

hospital.

Demonstration-Based Initialization of Reinforcement Learning Algorithms for Efficient Robotic Control

Kindred’s Sort product is a robotic system that operates in e-commerce distribution centers to sort and handle apparel and general merchandise. The deployed system is controlled through a combination of artificial intelligence and human-in-the-loop teleoperation. The proposed project involves applying techniques from artificial intelligence (specifically machine learning and reinforcement learning) to improve the ratio of automatic control to human control. The core hypothesis of the project is that historical data collected from human teleoperation of the robots performing object-grasping tasks can be used to train the robots to pick up items automatically. This task is a challenging research problem at the cutting edge of robotic control and AI, and it will be tackled with a combination of state-of-the-art academic research and internally-developed algorithms.

Exploration of Methods and Models to Achieve Multi-Document Comprehension in the Legal Domain

The project attempts to tackle an important challenge in Artificial Intelligence (AI), to give a machine an ability to comprehend multiple documents like humans do. These can do the redundant or preliminary reading-based research performed in many domains. The project aims to create a system which can read, understand, and answer queries and/or summarize multiple legal documents in a single shot. The project aligns with ROSS’s roadmap and vision – to supplement and enhance the quality and capacity of research tools, available at an average lawyer’s disposal and boost the time spent with their clients. Such a powerful system requires the capacity to understand queries made in natural language such as English. Hence the system will be developed using applicable novel state-of-the-art AI based Natural Language Processing (NLP) techniques. TO BE CONT’D

Indigenous Energy Financing Solutions

Energy poverty is an issue that impacts many First Nations communities in British Columbia. Despite significant appetite from communities to develop clean energy solutions, many do not benefit from the available financing programs. This research seeks to understand what barriers exist for First Nations communities in accessing energy financing, as well as to identify some best practices from other jurisdictions. The project will use this information to propose innovative community energy financing options that could be applied in the B.C. context.

Flexible Data Reader on Distributed File Systems for Training Deep Learning Algorithms

With the fast-growing size of machine learning datasets, it has become increasingly important to store them in a reliable and distributed manner. Large scale distributed file systems such as GFS, HDFS and Amazon S3 have the capability to store large scale of data reliably. However, these distributed file systems have an intrinsic shortcoming: they provide good read/write access guarantees only for large size files, and therefore cannot efficiently handle frequent read/write operations for large number of small files. In machine learning training protocols, the ability to shuffle data points within a dataset is crucial to avoid local minima and overfitting, which requires the data points to be accessed in a random manner, preferable efficiently. The main focus of this project is to find a way to store machine learning datasets on distributed file systems while maintaining a competitive randomly reading performance for shuffling data points. TO BE CONT’D

Automation and Evaluation of Rapid Deforestation Assessments of Small Farms

The worlds’ forests are being cleared at an alarming rate to make room for agriculture. Consequently, agricultural companies and those that finance them are increasingly required to demonstrate that the farms that they do business with were not recently deforested. Current methods for assessing deforestation involve using experts to analyze satellite images but this process can be cost prohibitive for small farms. The objective of this project is to develop and assess the accuracy of a new method that completely automates this process without the need for experts. Developing and assessing this method is expected to benefit Taking Root, the partner organization, to demonstrate that the farms that it does business with were not recently deforested at a much lower cost.

An order-reduced simulation framework for 3D multiscale geophysical electromagnetic problems

The goal of this research is to investigate the extension of upscaling and multiscale methods and their application to efficiently simulate (frequency-dependent and time-dependent) electromagnetic fields in geophysical scenarios that include metallic-cased boreholes and fractures filled with conductive/resistive fluids. Simulating this type of geophysical settings is quite challenging because they consider highly heterogeneous media and features at multiple spatial scales that require a very large mesh to be accurately represented. This results in a system of equations to be solved that often exceeds the limits of average computers. Thus, the key is to reduce the problem size but retain the accuracy of the electromagnetic responses, which is what upscaling and multiscale methods aim to do efficiently. The research-based solution proposed in this project is relevant to advance the prediction and simulation capabilities for geophysical electromagnetic problems in the context of Hydraulic fracturing. TO BE CONT’D

Étude comparative des effets maternels chez deux herbivores terrestres

Mon sujet de doctorat porte sur les capacités d’adaptation de grands herbivores face aux variations de leur environnement. Je m’intéresse particulièrement aux aspects liés aux soins maternels, tels que la date de naissance et la composition du lait maternel. L’hypothèse de base derrière mon projet stipule que les activités liées aux soins maternels devraient coïncider avec le pic de disponibilité de la végétation au printemps, pour que les jeunes et les femelles bénéficient de nourriture abondante. En effet, bien que les naissances aient lieu en majorité au printemps, peu d’études ont testé la capacité des femelles à s’ajuster à l’arrivée d’un printemps hâtif ou tardif au fil du temps, surtout en ce qui a trait à la composition du lait maternel.
Un stage dans le laboratoire du professeur Øystein Holand à la Norwegian University of Life Sciences, à Ås, en Norvège, serait une expérience très pertinente à mon avancement académique. Le professeur Holand travaille sur des questions très similaires aux miennes qu’il teste sur une autre espèce d’ongulé, le renne. La collaboration avec le professeur Holand permettrait d’établir un réseau de collaborateurs à l’international et de bénéficier de l’expertise de son groupe de recherche. “TO BE CONT’D”

The Contribution of Invertebrates to the Seasonal Diets of Walleye in Lake St. Joseph

The aim of this project is to better understand the onshore and offshore feeding habits and movement of walleye on Lake St. Joseph. There is special emphasis on answering the question: if, when and how much do Walleye rely on invertebrates in general, and Mayflies in particular. Walleye are an economically and ecologically significant sport fish and Mayflies are an important bioindicator of ecosystem health and potentially have an intricate predator prey relationship. To better understand these interactions, we will reconstruct the seasonal diet of walleye, across many age classes and determine their foraging habits both onshore and offshore. Biological information will be collected from the walleye to determine any specific foodweb interactions and transient movements that may exist. To conserve and protect ecological processes in an everchanging world understanding natural interactions has never been more important.