Automated AI-supported methane plume detection from satellite and aircraft images

GHGSat是一家加拿大公司，在从空间以高分辨率监测温室气体排放方面处于世界领先地位。它是利用卫星探测甲烷的先驱。GHGSat的产品和服务为世界各地的行业、政府和监管机构提供有用和可操作的排放信息。GHGSat的卫星和机载系统不断向地面发送大量图像，这些图像必须由专业操作人员进行单独分析，以确定甲烷排放区域。甲烷是一种温室气体，其影响是二氧化碳的84倍。GHGSat目前运行三（3）颗卫星和一个机载系统。在未来几年中，其他几颗卫星以及几个机载系统将投入运行。因此，随着这些新设备投入使用，收到的图像数量将大大增加。GHGSat项目的主要需求是一种工具，以支持其运营商确定潜在的甲烷排放区域，同时消除其他区域。为了提高其可靠性，该工具将把云量、风、太阳反射数据、工业场地的定位和过去排放的存在等信息纳入其分析。INO的数字视觉专家将参与该项目，以确定感兴趣的区域。随后，蒙特利尔理工学院将使用人工智能来提高检测决策的可靠性。该项目的目标是达到技术准备水平五级（TRL 5）。支持工具将被集成到GHGSat操作过程中，以优化图像处理时间，并更好地利用操作员的专业知识来处理困难的情况。此外，GHGSat希望能够减少其图像处理时间。GHGSat的专利卫星图像技术在世界上是独一无二的。这一项目的好处将直接影响我们作为世界领导者的地位和我们的业务。优化的运营、一致的分析、更快的交付给客户以及更好的排放控制将最终对环境产生有利的直接影响。

GHGSat is a Canadian company and a world leader in monitoring greenhouse gas emissions in high resolution from space. It is a pioneer in the detection of methane using satellites. GHGSat's products and services provide useful and actionable emissions information to industry, governments and regulators around the world. The satellites and the airborne system of GHGSat continuously send a large quantity of images to the ground which must be analyzed individually by specialized operators with the objective of identifying areas of methane emissions. Methane is a greenhouse gas whose impact is 84 times greater than CO2. GHGSat currently operates three (3) satellites and one airborne system. In the coming years, several other satellites will be in operation as well as several airborne systems. The number of images received will therefore increase significantly as these new equipments comes into service. GHGSat main need for the project is a tool to support its operators in identifying the regions of potential methane emissions, while eliminating the other regions. To improve its reliability, the tool would integrate into its analysis information such as cloud cover, wind, sun reflection data, the positioning of industrial sites and the presence of past emissions. INO's digital vision experts will be involved in the project to identify regions of interest. Subsequently, the École Polytechnique de Montréal will use artificial intelligence to increase the reliability of detection decisions. The project aims for a Technology Readiness Level five (TRL 5). The support tool will be integrated into GHGSat operations process in order to optimize images processing time and get a better use of the operators' expertise for the difficult cases. In addition, GHGSat hopes to be able to reduce its image processing time. GHGSat's patented satellite imagery technology is unique in the world. The benefits of such a project will have a direct impact on maintaining our position as a world leader and on our operations. Optimized operations, consistency of analysis, faster delivery to customers and better control of emissions will ultimately result in a favorable and direct impact on the environment.