Innovative Projects Realized

Applying Machine Learning to Develop Meaningful Rail Condition Indices

Rail transit and freight rail properties apply rail grinding to maintain rail condition and ensure satisfactory performance of rail infrastructure systems. The proposed research investigates and applies a variety of computationally intelligent algorithms to establish useful relationships between rail corrugation, noise generation, and vibration. These relationships will support more timely and effective rail grinding interventions. The algorithms will process real-world rail corrugation, noise, and vibration data collected from three rail transit properties in North America. The long-term research goal is the development of a generic and transferrable rail corrugation index, which will help rail maintenance practitioners determine when rail corrugation is likely to generate unacceptable noise and vibration. Consequently, the research directly supports rail and vehicle asset management programs, helps reduce noise irritation for passengers and citizens in the vicinity of rail transit lines, and improves ride quality.

Semi supervised object detection

Deep learning technology is a great tool to learn complex patterns and make prediction based on this learning. In order to get the most accurate predictions, one needs to train those neural networks on vast amount of labelled data. Labelling data is a time consuming and costly task. Using semi supervised learning, it should be possible to label a fraction of the dataset and let the neural network learn by itself on the rest of the, unlabelled, data, thus greatly reducing the overhead of using deep learning technology. This project aims at identifying and implementing the best possible semi supervised strategy for object detection.

Computer Vision R&D project – Autonomous Robot Assistant

The project’s main objective is to provide a robot with the capabilities to perform diverse helping tasks in the office, such as fetching objects, greeting colleagues, clients and newcomers. This will be a stimulating project involving tangible real-world application of multiple artificial intelligence approaches to robotics, including and focused on advanced computer vision algorithms, in the objective for the robot to navigate and recognize objects using a built-in camera.
This project will provide the company with a definite increase in its expertise in the commercial application of AI to robotics. Since Menya has already shown its skills in the space industry, this project may help get into new sectors in industry, but also in societal domains such as healthcare and education.

Power network transfer capability (Phase II – data error detection)

Hydro-Québec is a public utility that generates and distributes electricity. Despite selling most of its electricity in Québec, its most lucrative sales are in the neighboring markets. To ensure the best possible quality of service, the transmission system must remain stable, but to maximize profits, the company also wants to increase its transmission capacity to maximize energy exports. The transfer limit is now conservatively estimated based on a certain combination of simulated network configurations. This project aims to more accurately estimate the transfer limits of the electric grid and the uncertainty of these estimated limits. Recent advances in machine learning, especially in deep learning, in conjunction with more traditional algorithms used in computer science, have the potential to improve these estimates and therefore augment exports for Hydro-Québec.

The Development of Anti-gingivitis Probiotics Derived From the Human Oral Microbiome

The human oral cavity contains over 700 different bacterial species. In healthy people, these bacteria are living in harmony and not likely to cause diseases. However, sometimes this bacterial balance is disturbed as the oral pathogenic bacteria start to overgrow causing many oral implications such as halitosis, sore throat, dental caries and gingivitis. A promising solution to tackle this microbial population destabilization is the use of beneficial microbes called probiotics. Previous investigations showed that the oral commensal Streptococcus salivarius is an excellent candidate for the development of new probiotic treatments. This bacterium is human friendly and is one of the first microorganisms to colonise the babies few hours after birth. S. salivarius can produce unique molecules which can be used as molecular missiles to attack pathogenic bacteria and restore the microbial balance to the oral cavity.

With the Child in Mind – Brain Development and Best Interests Decisions

Everyday across Canada, judges from both the federal and provincial courts make ‘best interest’ decisions affecting the lives of thousands of children and their families. Infants and children whose cases come before the justice sytem are more likely to have experienced abuse, neglect, exposure to family violence, parenting impaired by addiction or mental illness, and substandard or unstable home environments. Lawyers, judges and child-serving professionals need to be educated in the science of early brain development so that they can maked informed, timely decisions that will have lasting positive effects on the health and development of children. This project aims to collaborate with family court judges and lawyers to identify their information and learning needs in the area of early brain developement. These needs will be adressed by creating and piloting learning resources that will allow them to meaningfully integrate the science of early child development into their daily practice.

Monitoring of turbine runner blade strains from indirect measurements using AI

Hydro-Québec has data acquisition systems for a multitude of sensors, some of which have been installed since almost 20 years in its electrical generation equipment (turbine-generator units – TGU). The collected data is primarily used to ensure that the information is adequate in the event of an equipment breakdown or for specific behavioral studies. Data from monitoring systems are little used in routine maintenance management activities, often due to lack of time and adequate and/or effective analysis methods.
Equipment maintenance is an important part of Hydro-Québec’s equipment management activities. The creation and maintenance of a surveillance system is a major investment for the company. With the development of machine learning analysis approaches, the goal is to provide operators with a clearer view of real-time asset status and predictions about their potential for use.

Short term Electrical Load Forecasting

Load forecasting is an essential activity for a company like Hydro-Québec. It is necessary for objectives as varied as the management of production or the management and maintenance of the electricity network. Any significant forecasting error can result in reliability issues, loss of opportunity, or additional costs to the business. On the other hand, a good prediction would allow Hydro-Québec to generate additional sales in neighbouring markets. With the deployment of its Advanced Measurement Infrastructure (AMI), Hydro-Québec now has a significant amount of new consumption data. This data can be used to improve demand forecasting, increasing reliability, decreasing expenses, and potentially generating new revenue.

Macroeconomic changes such as the decline of heavy industry, the recent changes in society (teleworking, variable rates, etc.) and in the future (transport electrification, behind the meter production, storage, smart grids, active role of the consumer, etc.) are current and incoming challenges for the parametric forecasting models such as those developed and currently used by HQ, since the load is more and more difficult to modelize with no clear physical phenomena and measures to explain it.

Active learning for visual detection on inspection robots

Robotics vehicles deployed at Hydro-Québec up to now are still mainly manually operated and human intervention is continuously required. The project aims to equip Hydro-Québec’s current and future fleet of inspection robots with autonomous inspection capabilities. The intern will leverage breakthroughs in artificial intelligence to enable robotic vehicles to realize real-time automated visual inspection of the company’s infrastructure and use a simply and securely deployable robotic vehicle to perform the company’s first fully autonomous power line components inspection mission. The large-scale deployment of autonomous inspection robots will have a major impact to help Hydro-Québec in asset management and gain in operational efficiency.

Solar Radiation Forecasting

The main duty of Hydro-Quebec is to respond efficiently to the energy demands of customers, in a safe way while remaining competitive in the markets as well. In a changing energy context, the production of solar photovoltaic energy represents a new challenge for Hydro-Quebec, which will have to integrate and to balance this intermittent resource to guarantee the reliability of the electricity grid. The objective of this project is to support Hydro-Quebec in the development of a future-oriented energy system by proposing innovative technological solutions notably for solar radiation forecasting. The project will focus on developing a state-of-the-art system using artificial intelligence algorithms to predict solar radiations 24 hours in the future from satellite images and numerical weather data.

GROKVIDEO

By combining the information contained in the visual, audio and text content of videos, it is possible to extract complex information about their content. It’s then possible to analyse a query from a search engine to find the video segments that best matches this query. During this project, the intern will be using state-of-the-art deep learning models to extract the best possible information from multi-source data and participate in the integration of these models in the Grokvideo search engine application. Increasing the quality of the information and the accuracy of the search engine will directly benefit the company as extracting the best possible information from video content is the core value of Grokvideo.

Brain Lesion Detection

Brain MRI scans are a critical component in the diagnosis of neurodegenerative disorders and their use will only increase in the following years. However, there is a wide diversity in terms of the image quality and resolution obtained across different sites and there is a need for robust methods that can handle such diversity. The goal of this project is to develop and validate the performance of state-of-the-art lesion detection methods for 3D brain MRIs.