Understanding memory consumption in a distributed data analytics system

Data analytics has emerged as an important driver for economic success in many businesses and in the public sector. In healthcare, using data analytics promises to reduce costs and improve patient diagnostics and treatment. This project is a collaboration with a Vancouver-based company PHEMI Systems that provides data analytics solutions for healthcare and other industries. Our goal is to understand how to make data analytics system used by PHEMI and many others more efficient in terms of consumed resources while at the same time meeting the needs of privacy.

Investigating How Teachers Learn and Customize Digital Classroom Tools II

Our proposed research investigates how K-12 teachers learn and customize digital classroom tools and learning management systems and how they share this information with each other. In particular, we will be working with our partner Microsoft to investigate the use and customization of the recently developed OneNote Class Notebooks software that is increasingly being used by teachers for various content delivery and content management tasks.

Ultra-low power wireless sensor node design for health monitoring use cases

In this project we address the problem of power consumption for wireless sensor nodes. This is where among different components of a sensor, RF transceivers consume a significant amount of power e.g. approximately 80%. Hence the main objective is this project is to tackle the power consumption problem at the RF transmitter, where we aim to reduce the power consumption to micro-watts of power, with minimal sacrifice in achievable data rate and by keeping the connectivity range within an acceptable radius.

Mapping the surface flow velocity of Minas Passage using RADAR data

This project will investigate the use of RADAR data to estimate the ocean surface velocity in regions of interest, specifically where tidal turbines will be deployed in Minas Passage, Bay of Fundy. The Fundy Ocean Research Center for Energy (FORCE) currently owns a single RADAR on the North side of the Minas Passage. Initial investigations have been done with this single RADAR; however, more intensive analysis must be done to reach the long-term goal of having a network of RADARs in the area.

Simulation and analysis of light scatter in head mounted display lens

Head-mounted display (HMD) lenses can include a high degree of scattering (ghosting) which reduces brightness and contrast, and is distracting to the user. This can directly impact the utility of the device, if for instance some of the display is illegible because of excess light scatter and blurring. In this research, we want to investigate ways to model sources of light scattering in a HMD with freeform lenses, using non-sequential raytracing software.

Deep Learning to assist requirement translation

Cutting edge techniques in artificial intelligence will be applied to extract semantic information from natural language and work towards building a system that can help engineers write clearer and less ambiguous requirements for complex systems. Models will be developed that are similar to current techniques used by popular translation tools, and will be adapted for paraphrasing and ambiguity reduction.

Advancing Out-of-Band Network Measurement for Multi-Hop Sensor Networks

This collaborative project with Rimeware will investigate out-of-band measurement approaches that can passively monitor the network traffic and provide rich detailed network information, e.g., latency, loss, route path, etc. The goal is to build a programmable system for accurate, generic, and robust network measurement. It includes two sub projects. In the first sub-project, we will research the sniffer deployment problem with lossy and correlated link models and develop a set of instructions that provide programmable interfaces for network administrators.

Evaluating the effectiveness of video as a medium of delivery and consumption of large quantity of digital information

The goal of the project is to evaluate the effectiveness of video as medium of delivery and consumption of large quantity of digital information. The team will create an automated way to generate videos for properties and rental information so as to reduce costs on manual video creation and build user profiles based on smart questions to present tailored information based on users’ needs in a video format. Through researching, implementing technologies, designing analytics tools and developing a prototype we hope to evaluate the effectiveness of dynamic video information delivery.

Fast and robust real-time precise point positioning - Year 2

The main outcome of this Mitacs-sponsored project will be a robust navigation software capable of providing accurate navigation solutions for commercial Unmanned Aerial Vehicles (UAVs). Such a software will further elevate the industrial competitiveness for the partner corporation, the Profound Positioning Inc. (PPI). After finishing this project, PPI will be able to offer more comprehensive embedded integrated UAV navigation products.

Advanced Learning for Automatic Object Detection and Tracking in UAV Imagery

Recently there has been a growing adoption of Unmanned Aerial Vehicles (UAVs) in many applications. Robust object detection and tracking algorithms are required in many UAV scenarios. However, existing object detection and tracking approaches are not specifically designed for UAV and therefore do not have satisfactory performance in UAV. In this MITACS cluster project, by working with the industrial partner, we will develop a set of improved object detection and tracking algorithms for UAV.