Innovative Projects Realized

Scalable Fault Detection, Diagnosis, and Situational Awareness for Modern High-volume Data Pipelines

Globally, industries are seeking to develop new products and services from the large-scale data sets they hold. As these systems move from prototypes into fully operational 24/7/265 commercial solutions additional services must be provided to detect and address system faults and failure as they arise. Within classical engineering plants, e.g., those of the telecommunications, petrochemical, transportation, etc. industries, these tasks are performed by fault detection and diagnosis (FDD) and situation awareness solutions. This Accelerate project will design and implement a prototype FDD and situation awareness solution for this emerging high-volume data pipeline industry. The solution will be tested and operationally assessed for two industry-held high-volume data pipelines, one in the telecommunications industry and one in the cyber-security industry.

Development of a supervised and transparent prediction model for predicting bond credit rating migrations in real-time for short to moderate time frame

The purpose of this research is to develop a supervised prediction model that will be used to predict whether a bond credit rating will migrate down to a lower credit rating in a short to moderate time frame. The problem will consist of the development and optimization of a bond credit-rating migration prediction model; the development of a framework that gives the model transparency and explain-ability lending weight to the credibility of the results; and the integration of the credit migration prediction into a portfolio allocation optimization for bond portfolios.

The model will provide insight into the short-term creditworthiness of corporate bonds that will be used to develop a platform for which the probability of credit rating migration will be available to a broad network of financial professionals in real-time. This can play a significant role in reducing the losses due to an un-expected bond credit rating downgrade.

Efficacy Assessment and Improvement of an Allergen Dispersion System in an Exposure Room

Red Maple Trials (Ottawa, ON) created a new facility where patients can be exposed to airborne allergens. In this facility, the patient’s symptoms (with or without medication intervention) can be monitored in a controlled manner. In the current project, airborne allergen (pollen and dust mites) concentration and size distribution will be monitored in time. It is expected that more homogeneous allergen concentration and spatial distribution in this exposure room will allow better competitiveness for the company to attract pharmaceutical companies. Nevertheless, allergy research regarding the approval of new medication (with better efficacy) can impact a significant portion of allergic patients in Canada and abroad.

Pore Pressure Prediction, Hydraulic Fracture Propagation and Huff-and-Puff Gas Injection in Multiple-Porosity Shale Reservoirs

Shale reservoirs store gigantic volumes of petroleum (oil and gas). However, because of the complex nature of the reservoir rock, it is difficult to recover the oil and/or gas stored in shales. Under normal conditions, it is possible to extract only as much as 10% of the resources in place, thus leaving behind a huge potential that promises to satisfy the energy needs of Canada for several decades. The proposed research project aims to provide a scientific methodology whose practical application will allow increasing the recoveries from shales, i.e., getting more oil and/or gas out of the reservoir. The project is therefore aligned with the interest of Canada for boosting its shale industry.

Machine Learning for Breath-Based Cancer Diagnosis

Non-invasive breath analysis has substantial potential for monitoring of a wide range of medical conditions and observation of overall health status. Breath testing is easy and painless; it can be done quickly and inexpensively, and can be repeated as often as needed, making it an attractive approach for screening or clinical diagnosis. In this work, we aim to leverage machine-learning principles to improve and validate the performance of a novel breath-based cancer screening tool.

Closing Skill Gaps in Under- and Unemployed Youth

Learning, an online learning platform, to underemployed or unemployed youth. Building on the work of a pilot project by CivicAction, the intern will plan and implement a randomized controlled trial to test factors driving uptake of online learning and its effects on employment outcomes. The intern and CivicAction will engage with Toronto-area employment services providers to provide various treatment intensities. Treatment groups range from a simple offering of a license to weekly reminders and curated playlists to encourage uptake or hands-on help in setting up online learning modules. Based on their pilot study, online learning has great potential to improve youth unemployment. This project will provide CivicAction with a rigorous evaluation of their online learning initiative. This evidence will inform future projects and support project expansion to larger populations.

A Clean Energy Plan for CleanBC: Informing the development of BC’s clean energy future

CleanBC is British Columbia’s recent climate strategy outlining how the province plans to transition from fossil fuels to a clean and renewable energy system. The plan contains targets that are meant to set a pathway towards a more sustainable future; however, Pembina Institute recently recommended that the government develop a clean energy plan to back up the CleanBC strategy by outlining more specifically how clean energy is going to be used to achieve the province’s clean energy targets (Riehl, Tam Wu & Kniewasser, 2019). Based on this recommendation, this project will look at ways to develop a provincial clean energy plan that works alongside CleanBC. The clean energy plan will include a creation of scenarios for transitioning the transportation system to lower carbon alternatives, such as hydrogen, in order to provide an example of how the provincial government can more clearly and transparently help reduce the use of fossil fuels. The project will help Pembina inform the provincial government on ways to achieve climate-related goals in a more efficient and rapid manner in a time when the damaging impacts of our consumption of energy on the climate are becoming increasingly unavoidable.

Improved Lateral Supports for Fresh Masonry Structures at Construction Site

Despite significant development of various construction materials, masonry is still considered as one of the most cost-effective materials. However, they are often vulnerable to wind-induced lateral loads caused during construction stage (within 1 to 2 days of construction) when the masonry is yet to achieve full strength. Temporary bracing systems are often used to support these structures at construction site to avoid loss of materials and injury to workers. However, design of these temporary bracings is relatively subjective and tend to be very conservative to maintain safety at workplace. The industry partner, Canada Masonry Design Centre, has an immediate urgency for improved efficiency of bracing of walls under construction which require a better understanding of early-age properties of masonry. This proposal aims at accurate characterization of early-age masonry and development of stringent design guidelines for bracings of newly constructed wall systems to eliminate the current conservative and uneco

Improving Efficiency and Robustness of Model-based Reinforcement Learning

Model-based reinforcement learning allows AI systems to learn and use predictive models of their environments to plan ahead, achieving tasks more efficiently. The proposed project aims to (i) develop methods for identifying when an uncertain and/or flawed model can be relied on to make plans, and when it cannot, and (ii) implement a method which allows an AI system to explore its environment exactly when exploration will be most useful for improving its model-based predictions and plans. Such methods for using models robustly, efficiently, and adaptively are promising for real-world applications of reinforcement learning which require systems to achieve tasks with limited data.

Data-driven Assessment of Suicide Risk for Treatment Seeking Population

As part of this proposal the intern will be working with DME to develop and examine the viability of data-driven point-of-care system for the assessment of suicide risk. DME is a Canadian start-up company in the business of developing cloud-based point-of-care monitoring systems for the management of psychiatric illnesses. DME has developed algorithms to diagnose and predict optimal treatment for major depression disorder and schizophrenia, and has been allowed patents describing its technology in Canada, the USA, and Australia. If found acceptable the algorithms

A Clinical Proteomic Test for Patient Care

There are 2.6 million Canadians with Chronic Obstructive Pulmonary Disease (COPD). COPD is a disease characterized by progressive loss of lung function that leads to shortness of breath, poor quality of life, reduced productivity, emergency visits, hospitalizations and mortality. The World Health Organization estimates that COPD will be the 3rd leading cause of death worldwide by 2030, accounting for more than 7 million deaths/year and 11,000 deaths/year in Canada. COPD patients frequently experience ‘lung attacks’, during which breathlessness, coughing, and sputum production dramatically increase, leading to urgent clinic visits, emergency admissions and hospitalizations. Lung attacks cost the Canadian health care system nearly $3 billion each year in direct expenditures. There are no blood tests to confirm a lung attack and because of this sometimes the doctors can make the wrong diagnosis, which leads to wrong therapies. In this project, we will develop a blood test to help doctors make the right diagnosis of lung attacks.

Identification of blood-based biomarkers predictive of pulmonary exacerbations in cystic fibrosis

Individuals with cystic fibrosis experience recurrent episodes of worsen in respiratory symptoms, termed pulmonary exacerbations (PEx). Early identification of individuals who are at elevated risk of PEx can improve their clinical outcomes and rescue their lung function. In this project, we will collaborate with the Prevention of Organ Failure Centre of Excellence (PROOF) to develop a simple blood test to predict the PEx in CF individuals. We will also evaluate the genetic influence on blood biomarker candidates during the project and refine the blood test based on these genetic variants. Our interactions with PROOF Centre will directly benefit their organization by providing proof-of-concept of their goal, to develop blood-based biomarkers and move towards personalized medicine.