Innovative Projects Realized

Assimilation of different data types with machine learning methods for resource classification

Classification of mineral resources based on geological confidence level is mandated for disclosure by all of the international codes and standards. However, the definition of the “confidence level” is vague in all of these codes mostly because of the fact that every deposit has its own characteristics. This research attempts to establish a method that utilizes advanced statistical methods to quantify this confidence level that transparent and practicable enough so that it can be adopted by the industry. A practical approach to uncertainty quantification task could provide the partner company which is one of the pioneer service company in this field, with an alternative reliable service to provide its clients.

Characterising and assessing seabird bycatch in expanding Arctic fisheries

Fisheries are an important industry in Canada, particularly in rural communities that depend on marine resources for sustainable economic opportunities. But all fishing imposes environmental costs. In the case of northern fisheries, one cost of concern is the incidental bycatch of seabirds by fisheries. Birds, attracted to bait and discards from vessels, can get ensnared in nets or hooked on longlines. The goal of this project will be to develop mitigation strategies, including applying best-practices to existing gear, modifying gear to reduce seabird entanglement, and develop strategies that minimize interactions between fishers and birds.

Information Asymmetry in a Fragmented Low-Latency World

This research aims at improving theoretical financial markets models in order to account for the presence of informed traders in a contemporary setting. Despite the fact that privileged information has always been present in various shape or form, theoretical models seemed to be able to account for such informational asymmetry and provide ways to measure it (e.g. Kyle (1985); Easley and O’Hara (1987)). Nevertheless, recent empirical research (Collin-Dufresne and Fos, 2015) has shown that these measures fail to detect the presence of informed trading. We thus wish to investigate the question of whether classic theoretical models are well equipped to capture the salient features of modern financial markets (which are called its “microstructure”). To do so, we will use a high-precision database which records every aspect of Canadian stock trading. TO BE CONT’D

Net Zero Energy Design Trade-offs and Excel Tool for Assessing Net Zero Potential for Multiunit Residential Buildings (MURBs)

Net zero energy building produces as much energy as it consumes, and it has a great potential for energy and carbon reductions. This research will study the design of net zero energy multiunit residential buildings (MURBs), as well as to create a simple Excel tool as an alternative to EnergyPlus for quick assessment. In the previous research, an EnergyPlus model for a real MURB was created and calibrated and strategy for this building to undergo dramatic energy reduction was studied. This project will continue to quantify the trade-offs involved in the design and then create the Excel tool. The partner organizations are among the leaders in pushing the residential development industry to higher energy performance standards. This project will provide comprehensive information relevant to the design of net zero condominiums. TO BE CONT’D

Smart building data analytics

RealTerm Energy is developing an AI-based solution for predictively optimizing HVAC (heating, ventilation, and air conditioning) systems in smart buildings. A substantial component of this solution is a data analyzer which will mine an enormous dataset to model the operational behaviours of the buildings in different locations around the world.
Based on enormous datasets collected from thousands of sensors during past years in several buildings, this project is aimed to extract patterns and build correlations among various metrics in order to model the behaviour of internal building environment. The ultimate goal is to detect shortcomings in current operational plans of the buildings, and then generate control policies in real-time that will minimize energy consumption according to real-time behaviours of each type of buildings, the architecture of the building, its geographical location, as well as outside weather condition

Analyzing Agriculture and Biodiversity Trade-offs in Three Sub-Saharan Countries at the Watershed Scale

With the onset of climate change and an ever-increasing human population, Sub-Saharan Africa faces a three-pronged sustainable development dilemma: how to increase food security while also conserving biodiversity and reducing social inequality. Given the importance of agriculture in Sub-Saharan economies, agricultural intensification and expansion will likely comprise the main course of action to address food security in most Sub-Saharan countries. Yet, agricultural expansion is the leading cause of biodiversity loss around the world and its effects on ecosystems will likely be exacerbated by those of climate change. Thus, from an ecological perspective, one of the greatest challenges will be to identify areas where trade-offs between agricultural development and biodiversity will be greatest. While such efforts are ongoing to ensure the protection of terrestrial biodiversity, less attention is being paid to trade-offs between agricultural development and freshwater ecosystems in Sub-Saharan Africa. TO BE CONT’D

Simulateur externe de poumon de nouveau-nés: une preuve de concept

Environ 7% des nouveau-nés sont atteints de syndromes de détresse respiratoire à la naissance et nécessitent des traitements exposés dans le programme de réanimation néonatale. Même en santé, les nourrissons ont besoin de traitements de qualité administrés par des soignants qualifiés.
CAE Healthcare (Montréal, Canada) forme ce personnel en santé grâce à des simulateurs avancés capables de reproduire des scénarios réels de nourrissons en santé ou en détresse. Pour compléter sa solution de simulateur de nouveau-nés Luna?, l’entreprise souhaiterait obtenir une preuve de concept lui permettant ensuite de développer un simulateur externe de poumon. Cependant, des solutions existent déjà sur le marché mais un compromis doit être trouvé entre la fiabilité aux paramètres physiologiques et le coût du produit fini. Le simulateur devra reproduire au mieux la mécanique respiratoire ainsi que les paramètres physiologiques de la respiration. Le simulateur convivial formera donc les soignants en les mettant en situation.

OPTIMIZATION OF AN AGRICULTURAL BASED-APPROACH FOR INCREASING STEARIDONIC ACID (SDA) ACCUMULATION IN THE PLANT BUGLOSSOIDES ARVENSIS USING CO-INOCULATION WITH PSEUDOMONAS SP. LBUM677

The objective of this project is to optimize a novel agricultural-based approach for producing high levels of stearidonic acid (SDA)-containing seed oil in the plant Buglossoides arvensis using co-cultivation with Pseudomonas p. LBUM677. Production of seed oil enriched in SDA, a dietary product possessing proven health benefits, is promising for use as animal and human diet supplements. SDA has been found in plant species and cultivars belonging to the Boraginaceae family that have significant potential for development into commercially viable crops. Production of SDA will be achieved through the use of proprietary plant cultivars provided by a commercial partner, Nature’s Crops International (NCI), in conjunction with proprietary bacterial microbial inoculants already available in our laboratory. Preliminary results have shown increased SDA yields per plant, and ultimately per hectare, using co-inoculation strategies with selected microbial inoculants. Indeed, the co-culture of plants with

selected microbial inoculants, such as plant-growth promoting rhizobacteria such as…………………………..

An emerging aquatic invasive species in Saskatchewan: distribution and ecology of the Prussian carp

Invasive aquatic species are one of the largest threats to aquatic ecosystems and cost Canadians around $343 million annually. Aquatic invasive species negatively impacts recreational use of freshwater and can decrease water quality, reduce native populations, and can cause fisheries to collapse. Prussian carp are an invasive species native to Asia that have recently been introduced to Alberta and has now expanded into Saskatchewan. To reduce the negative impact on Canadians, it is crucial to develop an effective management strategy, but to do that certain information is required. The distribution and diet of the species is two areas of information required to successfully manage Prussian carp in Saskatchewan and that is what this project will determine.

Optimal Execution of Trading Strategies

Financial Markets represent complex systems where the dynamics between market participants drive the evolution of prevailing prices. For this reason, it is important for fund managers to understand how these market impacts shape liquidity, transaction costs and equilibrium returns. To cope with the vast uncertainties, the subject of stochastic control provides key insights and critical solution methods to inform on optimal decisioning. Accordingly, the primary objective of this project will be to model the optimal execution of trading strategies which seek to minimize incurred transaction costs. Cast under a stochastic differential framework, this project will expand upon recognized tractable models by exploring new parameter estimates and revised numerical schemes for efficient computation. The success of this project will be measured by the ability of the framework to summarize critical economic insights that can motivate trading strategies used in practice.

Vision Based Tracking of Hands in Vehicles

Eye tracking has been a staple of partially automated vehicles for the last several years and forms a standard part of a driver monitoring package:
https://www.seeingmachines.com/industry-applications/automotive/

The next step is to understand if drivers are hovering over the steering wheel or pedals in preparation to take control. As part of the HumanDrive project: https://humandrive.co.uk/ we are collecting video for postprocessing of the hands and feet. With support from the School of Computing and Institute for Transport Studies the summer intern will leverage or develop a machine vision application to process the videos and track the hand position. A short report will be written that describes the approach and its reliability. This work would be some of the first of its kind and will prepare the intern for future machine vision research.

Earth Data Store – A one-stop shop for Earth Observation Data and Analytic Applications

UVIC and UBC, in collaboration with industry will be developing the Earth Data Store, to provide unparalleled access to standardized temporal and spatial earth observation data and develop industry-specific applications that will allow end-users to see how a region evolves over time using visual interactive maps and running deep learning algorithms. The Earth Data Store tackles the normalization problems of data, particularly with data generated from multiple sensors, which can consume up to 80% of the effort preparing for analysis. It is expected this project will support the growth of companies delivering big data/geospatial analytics products and information services, a global market estimated to be worth $10 B.