Innovative Projects Realized

Spectromètre à peignes de fréquences dans l’infrarouge moyen intégré sur puce optique

Les activités de recherche prévues au cours de mon séjour à l’University of South Australia consistent à développer un spectromètre à peignes de fréquences dans l’infrarouge moyen intégré sur une puce optique. Un spectromètre dont la plage spectrale d’opération s’étend jusqu’à 3-5 µm permet la mesure de spectres d’absorption de gaz. Bon nombre de molécules possèdent une signature spectrale autour de 3-5 µm et ne peuvent être caractérisées à l’aide des spectromètres actifs actuels à 1-2 µm. La présente proposition s’articule ainsi autour de la conception d’un montage optique comportant diverses composantes, dont une puce de verre fluoré qui agit comme le milieu de gain d’un laser et permet la génération d’impulsions laser de très courte durée formant un peigne de fréquence. À titre d’exemple, la télédétection, c’est-à-dire la quantification des émissions de gaz à effet de serre est une application où les besoins sont énormes et où la spectroscopie par peignes de fréquences s’avère une solution à fort impact à la fois pour l’industrie et la population.

Development of an internal Credit Rating model for unrated companies / Quantitative alignment of portfolio risk against a client’s risk appetite

The first part of this project involves the development of an internal credit rating model. Credit ratings, such as those provided by credit rating agencies like S&P and Moody’s, are an important tool that measure the credit risk associated with different companies. However, as the markets expand there are an increasing number of companies where investors would like to participate, but which are not rated by these agencies. Thus, the Toronto-Dominion Bank’s Wealth division would like to create a model that can assess a company’s credit rating and compute a similar score to that of the credit rating agencies. This internal rating system would allow TD Bank’s customers to better measure credit risk with their investments.
The second part of the project will attempt to improve the metric by which client risk appetite is quantified to better align the appetite level with the portfolio risk for each customer, thereby ensuring a client’s needs are effectively met.

Predicting the Ambulance Offload Delay Problem Using Decision Tree Analysis

When emergency departments (EDs) are congested and cannot accept incoming ambulance patients immediately, a common action is to let paramedics continue to provide patient care until an ED bed becomes available. This delay in transferring a patient from the ambulance to the ED is referred to as ambulance offload delay (AOD). AOD is a growing problem in Canada as the time to transfer an ambulance patient to an ED can be significant. This can negatively affect the ability of the ambulance service to respond to future calls and reduces the efficiency of the system. Using integrated historical data from a partnering hospital and partnering Emergency Medical Services (EMS) provider, a decision tree model will be developed to predict AOD of the system in a proposed period. The development of this prediction model can help ambulance service providers to be proactive with the potential AOD problem by activating interventions to prevent it from occurring.

Understanding the function of effectors’ structure-based function and protein-protein interactions from the amino acid sequence by computational calculations.

Identifying fungal effector proteins and understanding their function is of great importance in efforts to control losses to plant diseases. Modern high-throughput sequencing technologies have facilitated the availability of several fungal genomes and 1000s of transcriptomes. There is minimal consensus over the annotation and functionality of effector proteins. With the characterization of avirulence (Avr) genes, criteria for computational prediction of effector proteins are becoming more efficient. The proposed objectives are – 1) Effector’s 3D structure prediction, involving I-TASSER for iterative protein structure assembly, and ab initio protein folding using QUARK platform. 2) COACH based effectors-ligand binding affinities site prediction. 3) biological function annotation and predictions of effector molecules from its 3D structural model. 4) alignment-free bioinformatics approaches to identify a protein with structural and functional similarity to the Mlp37347 and Mlp124357. 5) To understand the 3D structure of Mlp37347-GAD1 and Mlp124357-PDI complexes from the sequence; and their structure-based function annotation.

Resilient Flywheel-Based Energy Storage Platform for Power Substations and Transportation Electrification

This project is aiming at the development of flywheel-based energy storage platform as integrated with power grid substation that will ensure sustainable and safe operation during normal and emergency conditions. It provides reduced operational costs with reduced GHG emissions in energy infrastructures. The design of a resilient energy storage platform is developed to be integrated with power substation and fast charging system for transportation electrification. The focus is to design a resilient energy storage platform, which includes battery and flywheel system, to be integrated with power substation to ensure stable and reliable power support to their customers. The integration of energy storage system will provide balance between the loads and supply while stabilizing the performance of substation. This requires dynamic storage systems to charge quickly, and to stabilize the dynamic behavior of the substation during charging / discharging cycles.

Assessing Statistical Bias in Credit Markets, an Application to SMEs

This research project aims to evaluate whether members of minority groups or women face higher barriers to access credit in the small and medium-sized enterprises credit market. The intern will analyze loan-level data provided by the business partner to evaluate whether these biases are detectable in the portfolio of SME loans of the business partner. Discrimination in credit allocation prevents efficient credit allocation, besides being demeaning for the individual subject to discrimination. Notably, the business partner is committed to providing fair access to credit to all its clients, independently from gender or minority membership. Unfortunately, biases may still influence the applicant screening methodology, due to the use of variables that are highly correlated with group characteristics. This research project aims thus to contribute to the effort of the business partner to insure that its clients face fair access to credit.

Modeling the effect of cloud conditions on incident spectra for concentrating photovoltaic (CPV) energy yield calculations

Concentrating Photovoltaic (CPV) systems use low-cost lenses or mirrors to concentrate sunlight onto the highest efficiency solar cells. They typically use 0.1% of the expensive materials used in traditional solar panels and achieve double the efficiency. As a result, CPV has emerged as a viable option for large-scale, low-carbon, electricity generation in sunny locations and will be viable in Canada within the next 5 years.

Predicting the electricity output, or energy yield, of CPV systems is a vital part of determining their profitability and the models used to do this currently rely on low resolution weather data that can introduce inaccuracies. This project will determine methods to treat low-resolution weather data, taking into account the effects of clouds, snow and other seasonal effects, in order to improve the accuracy of CPV energy yield models and therefore aid the wide-scale deployment of concentrating photovoltaics.

Hybrid Distributed Energy Resources for Net-Zero Energy Buildings

The main objective of the project is to develop hybrid distributed energy resource (HDER) systems to supply energy to net-zero energy commercial and residential buildings. This is expected to result in lower energy costs to consumers and utilities and in greater reliability of the grid. The HDERs will consist of solar panels, generators, and batteries. They will supply buildings with energy whenever possible and feed any excess energy to the grid. The grid will supply energy to the buildings whenever energy from the HDERs is insufficient. Over time, the net amount of energy supplied by the grid will be zero. The project will be a collaboration of researchers from Western University and personnel from the Centre Wellington Energy innovation, who will commercialize the HDERs under the product name of ONguard.

New Deformation Model for Realistic Real-Time Simulation of Soft Bodies in the Context of Virtual Surgery

OSSimTech develops and sells virtual reality (VR) surgical simulators. This proposal aims to improve the efficiency of the simulation of cutting and deforming organs. The current simulator used by the company is appropriate for a range of surgery scenarios, but for several scenarios, it is too long to compute. The company has plans to expand its portfolio of simulators to less powerful devices, potentially including portable devices such as tablets. In this context, the development of a low-cost simulation approach is of great interest for OSSimTech. Whether on high-end or low-end devices, the efficient simulation of soft bodies is complicated because of the real-time constraints. We will improve simulation models for the deformation of stiff soft bodies. Our goal is to improve the realism of real-time deformable tissues and organs in the context of VR surgery simulation.

Development and evaluation of a mental skills training program in youth hockey

This project seeks to develop a mental skills training program for kids in youth hockey. The first part of the project will involve reviewing the literature to find out what is knows about mental skills in kids. We are interested in seeing how kids can use mental skills in sports and hockey and if certain skills are more important depending on the age of the player. Once we have a good idea of what is known about mental skills in youth we will interview elite hockey players to find out what they think of mental skills training and what kinds of mental skills they think are valuable for kids. We also want to know if they thought mental skills training was available to them of if it would have been useful for them growing up as youth players. TO BE CONT’D

Développement d’éléments préfabriqués et de réparations pour les infrastructures avec les bétons fibrés à hautes et ultra-hautes performances- Phase 3

Les infrastructures en béton armé (ponts, routes, ports, canaux de navigation) ont un impact majeur sur la société. Les bénéfices économiques reliés à l’utilisation de ces infrastructures sont considérables, cependant ils sont contrebalancés par les coûts directs des travaux de remplacement ou de réhabilitations récurrentes. Le projet Développement d’éléments préfabriqués et de réparations pour les infrastructures avec les bétons fibrés à hautes et ultra-hautes performances (BFHP et BFUP) vise à développer des méthodes rapides et innovantes de construction, réhabilitation et renforcement durables des ouvrages avec les BFHP et BFUP qui présenteront une durabilité et une fiabilité nettement supérieures à celles obtenues avec les méthodes actuelles. D’une part, le projet permettra de développer de nouvelles applications structurales en BFHP et BFUP, soit des tabliers de ponts préfabriqués, des panneaux préfabriqués pour murs, des parapets avec coque préfabriquée. TO BE CONT’D

Order picking optimization at a pickup-van storage manufacturing warehouse

Ranger Design is a manufacturer of van storage units and accessories. Since order picking activities are considered to be an integral part of its warehouse operations and perceived to be conducive to significant improvements, re-engineering and optimization of the order picking process are in order. As such, the company is looking for an opportunity to reduce the travel time involved in order picking by re-engineering and optimizing the order picking process. This applied research project aims at developing an optimization model through which the most promising warehouse order picking operations in terms of travel time reduction are identified. The study will also address sensitivity analyses of input parameters. An implementation plan for the proposed order picking system will be presented.