Innovative Projects Realized

Natural Language Processing for automatically checking novelty of ideas

XLScout uses Natural Language Processing (NLP), Machine Learning (ML), and Innovation/Scientific principles to deliver actionable intelligence and accelerate innovation by analyzing large patent and research databases. The company is eliminating the pain of manually going through document and quickly providing relevant information to support data-driven strategic decisions. Presently XLScout hosts a data vault of over 130 million patents and 200+ million research publications occupying approximately 8TB of storage. Effectively searching these documents is time-consuming and it commonly requires advanced strategies that a novice searcher may not be familiar with. Moreover, as the database is so large, it is difficult to distill relevant information just by using keyword-based searches. XLScout already has different machine learning techniques for extracting information from these databases, and this project is seeking to make these systems smarter, efficient and scalable by utilizing state-of-the-art deep learning-based models.

Detours: Location intelligence for Inclusive mobility in disruptive sidewalk conditions

Temporary disruptions in the pedestrian environment, such as construction or snow, make it difficult for people with disabilities (PWDs) to reach destinations in their community. Cities struggle to communicate alternate routes that are accessible to everyone when these disruptions occur. Grounded in a context of data valorization and transfer for the development of smart and inclusive cities, the present project aims at addressing the challenges of the mobility of people with disabilities in a dynamic and changing environment (e.g., construction sites, social events, snow, etc.) to provide them with the information on accessible itineraries based on their personal profile, capabilities and preferences in outdoor environments in close collaboration with “Quartier de l’innovation de Montréal” (QI), the city of Quebec, as well as the “Réseau de transport de la Capitale” (RTC) and community partners such as ROP 03 (Regroupement des organismes de personnes handicapées de la région 03).

Structural testing of Basalt Fibre Reinforced Precast Concete Sandwich Panels

A new load-bearing precast concrete wall panel system has been proposed that uses composite materials instead of steel for reinforcement to reduce the level of heat loss through them and increase their R-Value. The panels will undergo destructive structural testing to determine how well the composite material system compares to a similar wall design that uses steel reinforcement. Bending tests, axial tests, and combined bending/axial tests will be preformed. The test results will allow for the development of a design aid for engineers that will tell them the bending strength of the wall under various axial forces. The partner organization, through the course of the internship, will learn the fabrication process for the walls and can then apply this process to future mass production. The resulting wall design aids can be presented to consulting engineers and customers that wish to use this system as promotional material.

Organizational Marketing Strategy Development-Standard Rail Corporation

Standard Rail is one of the leading participants in the Rail Service industry. We place the highest importance on ensuring employee safety and delivering quality customer service and pride ourselves on high-quality railcar and locomotive services across North America. This project would give the company the financial resources to branch out its service portfolio and market its service portfolio appropriately to build a niche reputation for itself in the market and help provide the necessary resources to support its marketing efforts.

Data Scientist / Machine Learning Engineer

1) Recommender System : With most modern services and products now being offered predominately online, it can be hard to get to know your customers. Unlike running a local store where you get to see each person, online businesses can struggle to understand their users’ expectations.
The project’s goal is to improve the customer experience by offering what they are looking for, thereby improving the conversion rates for retailers. To achieve these goals, the company plans to develop a Recommender system that predicts whether a particular user would prefer an item or not based on the user’s profile. Here the Recommender System draws on advances in machine learning to deliver personalized recommendations that suit each customer’s tastes and preferences across all your touchpoints.
2) Canada, a resource-rich country, garners a significant portion of the GDP from mining. With that said, most mines are still using legacy technology and facing the growing need to drive operations deeper underground.
Moreover, the industry faces volatile commodity prices and a decline in productivity despite continuous improvements in mining.
While artificial intelligence is still an emerging technology, it enables mining companies to become insight-driven enterprises that utilize data to pinpoint ores and lower costs. The project goal is to use machine learning to deliver value by instantly collecting data and deriving on-site insights that have the potential to vastly improve safety and streamline the workflow and develop an AI model that leverages the data captured from various sources and predicts the locations for ore mining.

Smart Textiles for Monitoring Aerobic Function using Artificial Intelligence

Physical activity is a crucial part of cardiovascular disease and prevention. Some of the most important clinical measurements relate to how effectively the body is able to consume and use oxygen to fuel muscles. However, these clinical measurements require complex technologies that make it only feasible in a laboratory environment. New technologies will be needed in the next 5-10 years for supporting remote monitoring. In this project, we propose to develop new fabric technologies with embedded sensors (“smart textiles”) that, paired with artificial intelligence, can monitor oxygen uptake continuously during exercise. These smart textiles will be evaluated in a clinical rehabilitation clinic to determine the feasibility of monitoring patients during prescribed exercise. Outcomes will produce new commercializable textile technologies for cardiovascular disease monitoring.

Information Extraction from Data Visualizations

Regulatory agencies publish several documents that outline the approval process of drugs. These contain valuable information on a drug’s safety, efficacy, etc. along with the feedback of reviewers from the agencies. Current technologies apply machine learning techniques to extract and categorize the unstructured text found in these documents. However, it does not accurately capture information from data visualizations such as tables, graphs, charts, etc. The data present in these elements are important for drug development teams to decide what clinical trial designs to use, which safety and efficacy data to gather for approval, etc. If solved, this can lead to quicker and more accurate decision-making by regulatory professionals, and in turn, result in faster drug approval and lowered drug development cost.

Development of a tele-health rehabilitation system to provide automatic assessment of patients’ performance, improve patients’ adherence and enable remote rehabilitation-(market and competitor analysis, regulatory & software design)

Poor physical recovery, especially in remote rehabilitation, is the problem that will be addressed in this project.
This project is Phase I of development of Fun-exercise module. Fun-exercise system uses gamification to boost patients’ adherence to their prescribed home-exercises. To achieve this goal, Fun-exercise will use mentally stimulating and customizable games paired with a set of wearable sensors to provide feedback to ensure activities are being done correctly. In phase I, patient study and conceptual design of the Fun-exercise module will be undertaken. Patient study determines and characterizes the target exercise and patient groups. This project will help the partner organizations to drive innovation and technology development in Canada. The work conducted by intern will inform the broader community about industry trends, the impact of different present and emerging technologies on various industries, as well as foster the growth of innovative technologies within the Canadian economy.

Evaluation of novel antiviral compounds against SARS-CoV-2

SARS-CoV-2 appears to be poised to become endemic, meaning that we can expect continuous risk of exposure to infection. Despite highly effective vaccines, people still become infected, and some people become severely ill. Antivirals that can reduce symptoms and decrease the levels and duration of infection can aid in reducing onward transmission of the virus. As part of a Canadian antiviral development pipeline led by Applied Pharmaceutical Innovation, candidate antivirals identified in high throughput screens will be evaluated for their activity against SARS-CoV-2 in live virus assays using cell culture models. Compounds with high antiviral activity will then be further investigated for their effectiveness using small animal models of SARS-CoV-2. Effectiveness in small animal models serves as proof-of-concept for further development, including moving the compounds into clinical trials. Identification and verification of the effectiveness of novel antivirals remains a key component of preparing a long-term strategy for dealing with SARS-CoV-2.

Anatomical fiducials and their relevance to stereotactic neurosurgery: quality control and improving surgical targeting

Ultra-high field Magnetic Resonance Imaging (UHF-MRI) allows for visualization of deep brain regions in exquisite detail, unlike any other imaging modality. In Canada, there are only a few MRI machines that can generate magnetic fields strong enough (7T or higher) to obtain UHF-MRI scans. The ability to visualize and locate deeper brain regions with millimetric accuracy is crucial to achieving optimal therapeutic outcomes during an invasive neurosurgical procedure called deep brain stimulation (DBS). DBS involves delivering electrical stimulation via electrodes to relieve symptoms of a wide variety of disorders like Parkinson’s disease, essential tremor, dystonia, depression, and addiction. Neurosurgeons often target deeper brain regions without seeing them due to limited access to UHF-MRI images.

The goal of my research is to develop a machine learning model (allowing computers to make decisions beyond their initial programming) to utilize the locations and distances between 32 points on brain images called anatomical fiducials (AFIDs) placed on UHF-MRI images to locate common DBS surgical targets with millimetric accuracy. I hypothesize that the 32 AFIDs survey deep brain anatomy enough to allow for predicting the location of multiple DBS targets.

Addressing the Discrimination experienced by Somali Canadians and Racialized LGBTQ Persons in Toronto

Addressing Racism in Toronto’’ is a one-year project to be conducted by Urban Alliance on Race Relations (UARR). The aim of this community-based research venture is to identify issues of access, equity and inclusion for two highly vulnerable and marginalized groups in Toronto: the Somali Canadian community and racialized LGBTQ persons who are homeless. Specifically, we will be hoping to uncover the barriers and challenges these communities face in relation to the following key areas: 1) employment, 2) health, 3) justice, 4) education, 5) housing, and 6) any other areas derived during the research process. We hope that by identifying the key social problems faced by the diverse communities that make up the Greater Toronto area, the Urban Alliance on Race Relations will be able to maximize its role and give voice to some of the most pertinent issues as related to race relations, access and human rights. The proposed research model, which will depend on the collaboration of the research group with various community stake holders, hopes to identify and review the current opportunities, needs, and challenges faced by the aforementioned communities.

Data and Visual Analytics for Prediction of Health Care Utilization Involving Musculoskeletal Disorders in BC

This study will establish the utilization rates of musculoskeletal services, for example, low-back pain, across healthcare systems in BC. To capture the utilization rates of musculoskeletal services, we can detect and establish the current frequency of these services related to the most recent healthcare expenditures. Knowing the utilization and expenditures would allow BC Chiropractor Association to update their allocation of funding with BC Ministry of Health to improve services while mitigating costs during the pandemic. Waves of COVID19 infections impact people’s health and could increase the burden of care and services by musculoskeletal practitioners. Moreover, the COVID19 pandemic increases the number of musculoskeletal disorders, especially lower back pain, arthritis, and other bone and spinal conditions of patients with moderate-to-severe cases in post-intensive care.