How Cerulean detects polluters at sea
Our tech partner AWS has released a new video that explains how SkyTruth’s Cerulean is changing the game in tracking illegal oil pollution at sea.
A little more than a year ago SkyTruth was honored to participate in the AWS re:Invent Hackathon for Good; one of only four nonprofits chosen to be a part of this annual event. We used that opportunity to draw on the brainpower of dozens of computer scientists and engineers to help jump start Cerulean — an effort to stop oil pollution at sea. At the same time we received an AWS Imagine Grant to propel us forward in 2020. In addition to financial support, AWS provided cloud credits and training.
Since then, our machine learning (ML) expert Jona Raphael has built a machine learning model that finds oil slicks at sea on radar satellite images. That ML model sits at the heart of an automated image handling and analysis pipeline in the AWS cloud that is running detections on hundreds of new images every day and steadily building a database of detected slicks.
With the brief video below, our partners at AWS are helping us share our work with the world and seek out those who can advance our goals — either by helping us develop the technology or by holding polluters accountable. In it, you’ll hear from Jona and SkyTruth President John Amos about why oil pollution at sea is a problem, and how satellite imagery, combined with machine learning techniques, is a promising solution in addressing a long-overlooked problem. Check it out and let us know what you think.