Projets novateurs réalisés

Image Processing for Digitalizing Engineering Drawings

Ce projet se concentrera sur la numérisation d’un type spécifique de dessin d’ingénierie en appliquant la technologie de traitement d’images. L’IPS a développé des technologies pour numériser différents types de dessins d’ingénierie, et identifié différents besoins technologiques et de développement de produits sur une feuille de route technologique. Ce projet est mis en place pour le traitement d’image des dessins isométriques de tuyauterie, et il est important pour construire des composants technologiques afin de répondre à certains aspects spécifiques de la numérisation du dessin (comme les lignes, les courbes et les formes).

Route optimization tool for vessels in ice-covered waters complying with carbon intensity index regulatory constraint

Route planning plays an integral part of each voyage in maritime operations. The selected route must be optimized in terms of economic factors and safety concerns. It also has to adhere to national and international regulations. Recently, the International Maritime Organization introduced a new regulatory instrument to promote the
decarbonization of the shipping industry. The regulatory instrument is called the carbon intensity index (CII). This project aims to solve route planning where all economic objectives are optimized while the operation adheres to the new CII regulation. The proposed method uses an Artificial Intelligence method called Reinforcement Learning
to formulate this issue. In the model, the vessel is an agent that explores the ice-covered environment to search for the best routes. A system of reward signals is defined to make sure the solutions achieve the optimality where their operation meets the CII constraint. This model will be tested in a simulation environment using a realistic voyage scenario in the Canadian Arctic.

Wind turbine interference suppression in HF-radar data

High-frequency surface wave radar (HFSWR) is recognized as one of the essential tools for remote sensing of the ocean surface. HFSWR received data contains valuable information that can be used for ocean wave forecasting, and since it provides real-time data, it can be applied for search and rescue operations, oil and pollution spills, and tsunami detection. When HFSWR is located close to the wind turbine farm, the spinning blades of wind turbines adversely affect the radar received data. In this project, we plan to investigate a real-time technique to estimate the turbine parameters and develop a method to mitigate the interference of the wind turbine in HFSWR received data for an arbitrary number of wind turbines.

Data aggregation, analytics and innovative ML application for Australian Clients-Koan Analytics Inc.

This research project is focused on data analytics for our Australian clients. One, focused on the data aggregation and exploration analytics for the State of Queensland, the other for the State of Western Australia. The interns will be working with cutting-edge machine learning and NLP approaches. This project will help increase the quality and speed of analytic runs we can offer to our clients.

IceCube – Shipborne Data Acquisition system for Arctic Autonomous Shipping

Le but de ce projet est de reprendre les techniques d’apprentissage automatique de pointe existantes et de les mettre en œuvre pour la classification des glaces dans les mers polaires. La classification de la glace joue un rôle crucial dans tout voyage de brise-glace. Un spécialiste de la glace à bord du brise-glace doit classifier tous les environnements de glace rencontrés. Ce processus est fastidieux et prend du temps. Ce projet vise à automatiser ce processus. En utilisant des réseaux de neurones de pointe, des images prises à bord de brise-glaces peuvent être utilisées pour classifier chaque pixel d’une image, offrant un contexte global et des informations sur l’environnement. Ces classifications de glace serviraient à générer une documentation importante pour les brise-glaces ainsi qu’à la formation de cartes de glace pour le Service canadien de la glace.

Hivenue AI Project

We are developing a P2P/AI marketplace, and we’ll integrate a matching process to let roommates match between them, or hosts to find the ideal tenant, based on different criteria, like lifestyle, interests, backgrounds, university, neighbourhood, space / time, reviews, market predictions, etc.

The candidate will do research development and will apply knowledge of advanced analytic algorithms and technologies (e.g. machine learning, deep learning) to deliver better predictions and/or intelligent automation.

The candidate needs to be familiar with programming languages related to the development of solutions using artificial intelligence in a computer vision context, and aware of deep learning platforms and/or libraries.

Machine Learning on Large Data Streams for CLS Operations

The CLS control system for the particle accelerator and beamlines, which are larger than the size of a football field, generates over 600,000 digital data streams to monitor and control the facility 24/7. The task of monitoring and reacting to these data streams is largely up to the Operators and other Egineers, Physicists and Scientists who manually check the data with the assistance of pre-set thresholds that generate alarms if limits are reached. Often once an alarm is triggered it is too late – the facility has dropped out of operations or equipment has been damaged. This project aims to take the first steps into a new paradigm of predictive maintenance and performance monitoring, by building modern tools to analyse the data for trending and correlations between data streams before an alarm limit is reached. The present method of manual human analysis is very time consuming and the CLS does not have the resources to retrospectively put all their data streams through algorythms that perform the same expert analysis provided by an expert human. This project aims assist the experts and provide a tool to analyse the huge data stream for possible correlations between the data streams that will lead to predicting common failure modes and prevent them from happening.

Model predictive controller and monitoring systems for Marine Autonomous Surface Ships (MASS)

For marine autonomous surface ships (MASS), a more acceptable operation is ‘constrained autonomous operation’ where the ship operates fully autonomously. Most of the existing systems have strictly defined operational constraints or limited available decision spaces; therefore, autonomous decisions are only allowed for some predefined scenarios. However, marine environment is dynamic; the environmental disturbances (wind, wave, current, ice), surrounding obstacles (ship, ice) can change quickly. In this project, we plan to develop an on-board navigation system using non-linear model predictive controllers (NMPCs) which will incorporate the path planning, collision avoidance, and tracking of the defined path and can deal with rapidly changing dynamic conditions in the sea. This cutting-edge research will help place Canada at the forefront of Marine Autonomous Surface Ship research. Our plan is to collaborate with various Canadian Marine Automation Vendors (e.g. Fleetway, Rockwell Automation, Siemens), Coast Guard etc. to fully harness the benefit of the proposed solution and improve the Canadian shipping industry in general.

Converting Fisheries Waste into a Sustainable Adsorbent for Arsenic Removal from Drinking Water

Arsenic (As) levels exceeding the acceptable level (10 ?g/L) in drinking water are detected in private, public, and school wells in many of Newfoundland’s small and remote communities. These wells are the main or sole sources of drinking water. This project aims to develop a sustainable solution to remove As from drinking water by diverting renewable fisheries waste from the landfill to manufacture an environmentally friendly adsorbent.

Seasonal Habitat Use and Sources of Environmental Stress of Fish in an Urbanized Coastal Inlet (the Port Moody Arm)

Burrard Inlet’s marine environment has been impacted by industrial and commercial activities and residential development for many decades. These coastlines provide critical habitat for various fish species. Port Moody Arm is one sub-basin of Burrard Inlet which has particular importance for Pacific salmon and forage fish. Industrial activity and development patterns in this area have changed in recent years. This project aims to collect data on fish populations and water quality conditions to compare against historic data and help determine current conditions. The results will benefit the City of Port Moody in many ways, including stakeholder collaboration, park outreach, and helping identify marine habitat restoration opportunities.

SSE_MichalisFamelis_GEARBOX_CharlesAntoineLord – Intelligence Comportementale de Groupe

Implement a squad based behavior in a chaotic environment.

The Spill Management and Response Tactics (SMART) toolbox for marine oil spill accidents

The Spill Management and Response Toolbox (SMART) can provide decision-makers and stakeholders with effective contingency planning by combining their expertise from multiple concerns such as response time, operational cost, or environmental impacts. The SMART system is designed to expedite the preparation phase, initiate response operations as soon as possible, reduce the workload and stress for decision-makers and stockholders in emergency planning and tactical meetings, and give on-site operators a clear understanding of their tasks. For example, field operators clearly know what to be picked up to ship A and B, and the detailed steps to follow once they arrive at the accident site. Reduced workload, less stress, and clear planning can help most people focus on just one or two tasks. Focused outbreaks can greatly reduce human errors and failures and reduce the system risks for a safe response. The toolbox has great potential as a product for customers in the government, industry, and academia.