Project M16

Empowering industries with real-time data to reduce methane emissions using satellite images and machine learning to detect and predict methane concentrations at specific coordinates

Comment

Inspiration

The inspiration for Project M16 stems from the Bhopal tragedy in India, where a gas leak went undetected due to the lack of proper infrastructure and real-time data. This disaster highlights the ongoing problem in many parts of the world, where industrial plants lack adequate monitoring systems, leading to catastrophic consequences. Our project focuses on providing real-time methane concentration data at specific locations, especially in areas with significant industrial activity, using satellite imagery and machine learning algorithms to track and predict methane emissions.

What it does

Project M16 imports satellite images from the Sentinel-5P satellite to monitor methane concentrations at specific geographic coordinates, typically those linked to oil and gas facilities. Using Google Earth Engine and real-time weather data, the project not only calculates methane concentrations but also predicts future levels based on environmental conditions like temperature and wind speed. This data is then displayed on a user-friendly platform, allowing industries to monitor and respond to methane emissions in real-time.

How we built it

The project was built in several parts:

Data Collection: Satellite imagery from the Sentinel-5P satellite was sourced via Google Cloud. This satellite provides the only open-source data for methane concentration and emissions. Analysis: Google Earth Engine API was used to analyze satellite images and calculate methane concentrations at specific latitudes and longitudes. Weather data was also integrated to enhance predictions. Prediction: A machine learning model was developed to predict methane concentrations based on environmental factors like temperature, wind speed, and atmospheric pressure. Reporting & Visualization: We built a front-end platform using React, and Django for the backend, to visualize the data in real-time.

Challenges we ran into

One significant challenge was figuring out how to access satellite data. Initially, we encountered roadblocks with closed platforms, but we eventually found that Google Cloud hosts the Sentinel-5P satellite data, which enabled us to proceed. Another challenge was selecting the right machine learning model that could incorporate a variety of test cases, including environmental variables.

Accomplishments that we're proud of

We are particularly proud of integrating satellite images into our project and successfully calculating methane concentrations at specific locations. Despite the challenges, we developed a full-stack application in a short time, leveraging machine learning models and real-time data to make methane emission monitoring accessible and actionable.

What we learned

We learned how to collaborate as a team and integrate machine learning models with satellite data. On the technical side, we gained experience working with Google Cloud, satellite imagery, and the Google Earth Engine API. We also improved our understanding of full-stack development using React for the front end and Django for the backend.

What's next for Project M16

Next, we aim to expand our capabilities by writing a proposal to the European Space Agency (ESA) to gain access to Sentinel-5P and its sister satellites, enabling us to monitor methane emissions on an hourly basis rather than daily. This would provide even more precise real-time data, helping industries mitigate methane emissions more effectively. We also plan to explore further collaboration with Google’s GPS and satellite ecosystem to refine our predictions and gain additional insights into methane concentrations globally.

Share this project:

Updates