Grants and Contributions

Farmers and agronomists spend substantial time travelling to each field to monitor plant health, disease, and the environment. AIRM Consulting is developing an innovative artificial intelligence-driven device for sensor-based 3D autonomous ground-truthing that integrates light detection and ranging (lidar) and multi-spectral sensors. Computer vision and machine learning are used to provide a real-time solution that is scalable, mobile, and optimized leading to novel quantitative data and crop monitoring, driving adoption of precision agriculture-based technologies with applications to decision support and insurance. This new technological solution will overcome hurdles pertaining to cost, scale/range, image quality, computational intensity and seamless interoperability.

This project aims to develop innovative approaches using Earth observation data for modeling and mapping floods, the most frequent and damaging natural disaster in Canada. By integrating optical and radar satellite imagery, this project will generate near-real-time flood boundaries and flood maps. These results will help decision-makers to have a global and current vision of the risk in order to take preventive action and optimize interventions, and for insurance companies to better evaluate risks for their clients.

Operating on an a Cloud infrastructure, this product will provide value adders the opportunity to offer their own applications as a service to users. This project will improve geospatial data access and usability for new users including non-experts and/or other non-traditional user groups engaged with application developers in a variety of application domains. Based on standards and best practices this project promotes system interoperability and the future creation of a global network of Earth Observation data resources.

Government authorities and data scientists do not have the ability to directly request ground truth observations from citizens to assess the validity of space-based Earth observation (SBEO) algorithms, and citizens do not have the ability to submit such observations directly. The opportunity is to combine these two observational approaches in a completely new method of environmental and disaster assessment using SBEO techniques and imagery. This will help close a communication gap and allow scientists and government officials to collaborate with citizens to validate and enhance the satellite image processing algorithms directly and in near real time, to determine the ground truth.

The objective of this project is to create an automated data processing pipeline and provide access to this data through a web visualisation platform that can be used for monitoring infrastructure extending across thousands of kilometers, while still highlighting displacement areas as small as tens of meters. 3vG will develop more efficient techniques to capture displacement features at all scales and displacement feature highlighting and filtering to allow end users to quickly focus in on areas of critical importance. The end result will be higher quality, lighter weight, and easier to use data products for network wide infrastructure monitoring.

The project is a wildfire management system developed in collaboration with INO, using requirements from SOPFEU and Natural Resources Canada to accomplish near-real-time, high-resolution thermal mapping of active wildfires using satellite/aerial imagery. Using pseudo image datasets, Lux Aerobot will be able to train deep learning computer vision models designed specifically to extract hotspots on a pixel level. To address potential telecommunications bandwidth limitations, the model will also be further adapted to be able to operate directly on-board satellite platforms, subject to the various constraints such as finite power supply, heat dissipation and limited computing power.

This project aims to develop a technology that can provide a reliable estimation of snow water equivalent (SWE) for monitoring and forecasting potential snowmelt flood events in Canada. This will be done through utilizing RADARSAT Constellation Mission compact polarimetric (RCM CP) data, by using innovative deep learning-based classification and segmentation techniques to map/segment snow coverage extent. A.U.G. Signals will also perform accurate snow depth and snow water equivalent estimations. The final result will help local hydrologic and disaster management services to determine the risks of snowmelt-induced flooding at a given location.

This project will implement a proof-of-concept and a feasibility validation of the ability to use space-based Earth observation (SBEO) data to quantify both the amount of phosphorus phytoextracted and carbon sequestered for a set of varied field trials carried out in the Lake Winnipeg basin. The project will set the foundations for a follow-on project to perform advanced application development to further demonstrate and refine the capability in the context of broader scope and reliability as a beta service in preparation for full commercialization and operation of upcoming global multi-spectral monitoring SBEO missions.

GHGSat will develop an innovative capability for validated dynamic three-dimensional greenhouse gas emissions estimation on a global scale using data from its own satellites and third-party satellites. Deliverables will include the implementation of a mathematical model for global methane emission rate estimation with components for the three-dimensional simulation of sources and sinks as well as chemical transport and atmospheric propagation. Once the emission rate estimator is validated, it can be adapted for operational solutions to address the requirements of regulators and end users in energy production, resource extraction, power generation, agriculture, waste management, and environmental sustainability.

The goal of this project is to develop a software framework that can be used by a variety of stakeholders (agriculture, finance, government and others) to develop, visualize, and integrate historical and operational evapotranspiration (ET) estimates. For this proof of concept, a software framework will be developed, tested, and applied by building a 10 year archive of crop ET using a multi-model ensemble. This software framework will provide a workflow that will allow users to upload files or draw polygons in order to get user-defined spatial and temporal summaries.