Innovative Projects Realized

Mixed Ni-Fe-oxide cathodes for hydrogen production by water electrolysis

Traditional energy sources, such as fossil fuels, are rapidly being depleted. Furthermore, CO2, one of the byproducts of burning these fuels, increases environmental risk such as climate change. Therefore, novel renewable energy paths are being developed to reduce the dependency on fossil fuels. Hydrogen, a clean energy carrier, is considered to be the best alternative to the current fossil fuel energy economy. Hydrogen can be produced by electrolysis of water applying energy which split the water molecule and forms hydrogen on the cathode surface. Taking into account this information, the goal of this project is to develop new materials for the cathode, in this case particularly, nickel-iron oxides, in other to achieve a higher hydrogen production by water electrolysis at a lower cost, since the materials currently use are very expensive. In addition, these new cathodes are expected to have a better long-term stability and to perform well in the alkaline and acidic media, and seawater.

Longitudinal Weak Labeling for Lung Cancer Prognosis and Treatment Response Prediction

This project aims at evaluating whether recent results in deep learning models, trained to exploit weak labels (Hwang, 2016) can serve to extract meaningful lesion localizations from image-level labels, either from individual scans or given a (longitudinal) sequence thereof. To this end, we will scale up existing models that have been shown to work on 2D images to a 3D context, studying labeling performance as the dataset size grows. If successful, this work will assert the usefulness of DCNNs to provide a general modeling framework to integrate imaging with other clinical patient data into a predictive system that could help support clinical decisions and ultimately improve patient care. The proposed research project fits within the partner’s scientific roadmap, which is to develop deep learning models suitable to processing clinical data that arises in a sequential fashion at the patient level (longitudinal data), wherein the set of available clinical modalities can be highly variable (heteromodality). The industrial partner has an existing team of full-time researchers dedicated to studying these questions; the intern will attack complementary questions with the help of the team. TO BE CONT.

Non-destructive Thermographic Stress Analysis of a New Composite Plate for Femur Fracture Fixation

The present aim of this study is to use an infrared thermography technique to non-destructively

measure the three-dimensional surface stress field in a synthetic femur fracture model repaired

with a new composite plate vs a clinically-used metal plate. To this end, there are three main

phases of this study. First, the infrared thermography system will be calibrated, which is needed

because the composite plate is made of several layers of woven material which have differing

material properties resulting in differing thermographic properties. Second, peak stress on the

new composite plate and the host femur will be identified as sites for potential mechanical

failure, thereby allowing the optimal repair method to be determined. Third, the mechanical

performance will be compared of a standard metal fracture plate vs. the new protoype composite

plate made from a polymer-based composite material.

Technology and Tools for Quantitative Neurodiagnostics Using Ultra-High Resolution Magnetic Resonance Imaging

The project aims to translate developments in ultra-sensitive MRI sensors to a clinically-relevant setting. To create high-sensitivity sensors for better images, we aim to create a tight-fitting system which places the sensors—akin to antennas—closer to the brain. This will improve the quality of the signals that we can extract from the brain, and allow us to use these improvements to capture images that have higher resolution and better contrast. Using this imaging improvement, we aim to then create a large normative dataset of grey matter thicknesses. This dataset will tell us what is “normal” for thickness in every part of the brain, and let us capture differences accurately and sensitivity. Ultimately, we aim to become sensitive to even subtle changes in grey matter loss, which may permit us to detect certain neurodegenerative diseases earlier, allowing us to treat them better.

Bioprinting an implantable knee meniscus

Each knee contains two menisci, crescent-shaped pieces of cartilage that play a crucial role in absorbing shock and providing nutrition to the joint. The meniscus is one of the most commonly damaged areas of the knee, unfortunately the body cannot easily repair meniscal injuries, leaving patients with reduced mobility and severe pain. Surgical removal of all or part of the damaged meniscus relieves acute pain, but often leads to osteoarthritis (OA) of the knee. Meniscus implants are a potential solution, however none of the currently available replacements prevent the development of OA.
Bioprinting is the fabrication of 3D structures from biocompatible materials. In this project, we will use novel bioprinting methods to build biocompatible meniscus-like tissues and validate their suitability for surgical implantation into the knee. Success of this project will enable Aspect (partner organisation) to progress to the next phase of meniscus implant development by testing printed tissues in animal models.

Learning representations through stochastic gradient descent by minimizing the cross-validation error

Representations are fundamental to Artificial Intelligence. Typically, the performance of a learning system depends on its data representation. These data representations are usually hand-engineered based on some prior domain knowledge regarding the task. More recently, the trend is to learn these representations through deep neural networks as these can produce significant performance improvements over hand-engineered data representations. Learning representations reduces the human labour involved in any system design, and this allows in scaling of a learning system for difficult problems. In this project, we propose to design a new incremental learning algorithm, called crossprop, for learning representations based on prior learning experiences. Specifically, the algorithm considers the influences of all the past weights while minimizing the current squared error, and uses this gradient for incrementally learning the weights in a neural network. This algorithm is called crossprop because it learns to shape the weights in a neural network through leave-one-out cross-validation procedure.

High power all-fiber Raman laser at 1.65 ?m

Fiber lasers have become the fastest-growing laser with a projected worldwide revenue up to $1.41 billion in 2017. In particular, fiber lasers at 1.65 ?m have drawn increasing attention with potential applications in chemical sensing, LIDAR and spectroscopy. All-fiber Raman lasing technology is a promising and efficient technology to achieve high power lasing at 1.65 ?m. However, there are limited all-fiber high power sources at 1.65 ?m that are commercially available. In this project, we will unite the expertise in fiber optics and lasers of the Photonic Systems Group at McGill University with those of O/E Land Inc. to develop a low cost and compact all-fiber Raman laser with a high-power output. Such a laser product addresses a gap in what is commercially available, and will add to their technology portfolio and give them a competitive advantage for different applications in spectroscopy, sensing and instrumentation.

Development and validation of training load metrics and models for predicting athletic performance

This series of projects will provide coaches and sport scientists with a greater understanding of the relationship between training and performance. While there are several methods for monitoring how much and how hard athletes train, how these can be best used to predict future performance is still in question. The sports of rowing and middle-distance running involve similar race demands, that being a full effort over 5-10 minutes. That said, the impact an athlete endures training for each is quite different and this can result in a limitation to time spent training in running relative to rowing. We will both investigate current methods of monitoring training for their use in making these predictions, as well as develop new approaches with the aim of even better predicting athlete performance resulting from the different training approaches taken by a coach.

Scale Up of the Circulating Fluizied Bed Bioreactor for Municipal Wastewater Treatment

The project will focus on the development and installation of a modification to convert existing biological wastewater treatment systems (particularly, activated sludge and similar processes) to circulating fluidized bed bioreactors (CFBBR). The CFBBR has already been proven on the lab and pilot scale to have higher nutrient removal efficiencies and greater handling of high volumetric loadings. Following the installation, the enhanced removal efficiencies will be tested. The system’s ability to handle high volumetric loadings will also be tested by monitoring the effluent quality during wet weather flows or by increasing the flow with clean water to simulate wet weather flows. As the CFBBR hasn’t been tested on this scale before, the analysis of the system following installation will also include troubleshooting for any unforeseen issues. Given the results the from past studies, the modification will likely enhance the treatment capacity. However, it is mechanical/operational issues that will need to be addressed.

Cooking Technology and Regional Identity during the Shang Dynasty

This project explores regional differences during the Shang dynasty (1600-1045BCE) in China by investigating cooking practices and cooking technology. Since cuisine is intimately connected to local culture, researching different approaches to cooking in the archaeological record can help us understand how different regions developed their own culinary traditions and identities even under the same political rulership. I explore this using three sites during the Shang dynasty – two from Northern China, Zhengzhou and Yinxu, and one in the south, Panlongcheng. Panlongcheng was a military outpost in the south for Zhengzhou, which means its settlers were Shang migrants that became exposed to different resources, soils and groups that likely impacted their cooking and cooking technology. This research will not only help develop our understanding of differing Shang dynasty populations, but also our understanding of the complex processes involved in the development of regional identity and local cultures in the ancient past.

Hardware-in-Loop Simulation and Test Facility for Re-Purposed Battery Energy Storage Systems

Hybrid vehicles have been in use for well over a decade now and have gained momentum in popularity

ever since. When these vehicles reach their end of life, their battery packs often have a considerable

amount of residual life. Although their state-of-health may not be suitable for vehicular applications any

longer, they can be re-used in a different setting where they can store and provide energy for remote and

off-grid loads such as small communities in the North. This project aims to develop an advanced

computer modeling and test platform to investigate various options for using re-purposed battery packs

for grid storage applications.

Real-Time Signal Optimization and Emissions Estimation Using Big Data Sources

In recent years technological developments have created a new paradigm where data can be obtained easily and with less effort than in the past. This shift is often called “Big Data”, and its effects can be seen as in many different fields. This proposal follows the same vein, and focusses on taking advantage of the increasing prevalence of connected devices. Modern devices broadcast unique addresses as they attempt to connect to WiFi or Bluetooth networks, and these addresses can be used to obtain estimates of traffic parameters such as volume, travel time, turning movements, and even the emissions generated by vehicles. These parameters can then be used to optimize the traffic system by changing signal timings and then collect feedback on the results of these changes. This project will take advantage of the wide array of resources and support available at WHUT to develop new and innovative technologies that use these data sources.