Recurrent Deep Architectures for Modeling Time Series Data

Deep learning is currently the dominant machine learning technique as a result of state of the art performance in vision (Russakovsky, et al., 2015), speech (Amodei, et al., 2015) and natural language processing (Vinyals et al., 2015). The improvement in performance of these models is attributed to the availability of large datasets for training the models as well as software & hardware improvements that help accelerate the training process. Recurrent Neural Networks (RNNs) are one of the most powerful and popular frameworks for modeling sequential data such as speech and text.

Intelligent Residential Energy Management Utility Controller

To research, design, and develop a network communication and control modules that integrate any residential HAVC control system with a utility energy management user interface. Developed signal modulation scheme will be implemented on development testing board. Device will network with all utilities for gas, water, and electricity.

Study of the Catalytic Effect of MFD on CuFeS2 Leaching

Bio-heap-leaching is a hydrometallurgical process used to process low grade chalcopyrite ore as the cost of alternative routes of processing and refining are not economically viable. The limitation however of the heap leaching process is the long time it takes to leach the metal and the low total recovery that can be achieved. As heap leaching being a large scale atmospheric leaching process, neither temperature nor pressure can be changed.

Statistical and Physiological Beat Modelling of Seismocardiogram Signal

"Seismocardiogram (SCG) is a signal that is captured by placing an accelerometer on the human chest. This signal captures very important timing information such as opening and closing of the heart valves. In addition to these timing information, the non-invasive nature of this signal makes it an attractive solution for remote monitoring of patients with heart conditions.
The morphology of SCG signal changes depending on different types of heart conditions and diseases. A mathematical model represents the morphology of a signal in terms of certain parameters.

Oil and lipid improvements in field pea to develop a non-traditional oilseedcrop

It has been noted in recent studies that provided an increase in the lipid content of the field pea (Pisum Sativum L.) through genetic manipulation, it can be used as a viable commercial alternative to conventional oilseed crops, which include canola and soybean. Genetic transformants with high lipid content can be created in the McGill University laboratories but its commercial viability needs to be tested with an industry partner.

Utilization of Supersolidus Liquid Phase Sintering (SLPS) in Metal Injection Molding (MIM) for Superalloys in aerospace applications

Powder metallurgy uses metal powders to produce parts of varying complexity. The processes can generally be divided in two big steps. The first is to form the powder into the required shape. This is generally done by pressing or molding the powder. The second step is to consolidate the powder into a solid piece of metal. This is done by heating the formed powder just below its melting temperature. At this point the metal particles will slowly coalesce into a uniform metal structure.

Controlling microbial processes in fracing fluids

Unconventional gas reservoirs are a great energy resource in the province of BC and Alberta and thus for Canada in general. Extracting this resource is not as straight forward as conventional gas reservoirs and requires hydraulic fracturing, also known as fracing or fracking, which has recently become a controversial topic in the public eye. This research project will examine water and fluids used in and returning from frac operations with respect to its geochemistry and bacterial populations.

Segment-based Fleet Management System for Semi Real-Time Analysis

Traditional fleet management systems suffer from zone-based aggregation. i.e. they use aggregate zone-level road and terrain characteristics to estimate trip performance indicators. In addition, most of these systems require intermediate ad-hoc staging tables to generate trip performance reports. To address these limitations, this research aims at developing a segment-based fleet management system for near real-time analysis. The proposed approach uses static segmentation to associate road and terrain characteristics with each segment of the road network.

Application of Variable Speed Drives for Improved Grinding Energy Efficiency

The research program is aimed at developing operating systems that enable grinding mill speeds to be controlled in responses to variations in ore properties. Although there are studies that show speed control can improve productivity and significantly reduce energy requirements, mines presently used fixed speed systems for their ball and tower mills. With development of new variable speed drive systems that can retrofitted to the fixed speed systems, there is an opportunity for mines to introduce the technology for their operation.

Optimization and validation of carbon nanofiber catalyst supports in fuel cell stack

Motivated by the urgent need for clean and sustainable source of energy we propose to develop structurally and chemically controllable fuel cell catalyst layers based on ultrafine nanocomposite carbon fibre catalyst support. Manufacturing parameters will be controlled and optimized to investigate the effect of microstructure on key performance factors. Ultimately, the knowledge gained from this study will pave the way to building more efficient fuel cells. Current phase of the project involves validating our design by in-situ testing.