The use of polyethylene (PE) plastic pipes for transporting gas and water has increased over the last few decades. The quality assessment and overall safety of the PE pipe networks have been of high priority for the distribution companies. The existing technology is somewhat complicated to use quickly and effectively by less experienced personnel. The proposed research study goal is to revise and evaluate the newest technological solutions for joint testing and propose a methodology that can address industry needs.
Avcorp Industries provides the world’s leading aircraft manufacturers with supply chain solutions and repair support. Yield optimization, predictive maintenance, and equipment calibration are needs that are widespread throughout the manufacturing industry. The root cause of failures in product testing is often difficult to determine particularly when the failure signals are sparse relative to the available background data. Compounding the problem, the process must meet a variety of specifications for multiple customers simultaneously.
Over 70% of tasks in manufacturing are still manual and because of this over 75% of the variation in manufacturing comes from human beings. Human errors were the major driver behind the $22.1 billion in vehicle recalls in 2016. Currently when plant operators want to gain an understanding of their manual processes, they send out their highly paid industrial engineers to run time studies. These studies produce highly biased and inaccurate data that provides minimal value to the manufacturing teams.
SOTI Inc. is a Canadian company providing control and management for mobile devices. Insight Agent is a product made by SOTI to help collect various battery specifications such as battery level, voltage, current and other metrics from mobile devices. This research project aims to use a machine learning and neural networks framework to predict the state-of-health for batteries in mobile devices. The intern will use the metrics collected by SOTI Insight Agent to derive formulas to calculate the key performance indicators (KPIs) of the battery system.
This project aims build an autonomous underwriting system that can provide debt financing to small and medium enterprises (SMEs) in the technology sector without human intervention. The research will explore mathematical methods and key factors that are unique to evaluating creditworthiness of those SMEs. The result will be an underwriting system that incorporates statistical methods and machine learning algorithms to perform prediction with higher accuracy than what humans can achieve.
An increasing number of wearable devices collect and use physiological information to track physical and mental health on a daily basis. While large-scale research initiatives allow an unprecedented amount of data to be collected, biosignal analysis techniques have yet to catch up. Indeed, analysis tools designed by hand based on small datasets available in traditional research settings are still widely used.
Ubisoft records the interaction between its customers and its servers in large execution logs (also called traces). Any failure of the system is thus recorded therein. However, the considerable size of these logs considerably hinders their effective use by analysts and developers. We propose an automated method to detect failing executions, and furthermore to characterize the features that are common to clusters of failing instances. The approach will be based an machine learning algorithms, and will produce clusters of failing traces with common features.
Ciena is a Canadian company leader in engineering and manufacturing networking systems and devices. The company has around 5,000 operable products in its portfolio. The vast majority of Ciena products generate logs during the boot up and the mission mode operations from the various tasks running on their real time operating systems. The company wants thus to increase its software’s capabilities in order to be able to collect any type of log data generated in the production site and linked to other external information to extract actionable knowledge.
This project seeks to explore the use of a class of artificial intelligence algorithms called reinforcement learning for the purpose of aiding the training of new pilots. In the process, we seek to “teach” an algorithm how to fly an aircraft by exposing the AI pilot to a virtual environment and providing it with flight data and a goal. Alternatively, the algorithm could learn by observing human pilots.
We’d like to address the issue of 3D reconstruction from 2D images. This means developing a machine learning algorithm that can take a regular photo as an input and generate a full 3-dimensional reconstruction of the contents of the photo. Such technology can be used creatively or to help the coming generation of robots better understand their surroundings.