Ocean Cleanup’s Efforts to Remove Ninety Percent of Ocean Plastic by 2040 by Leveraging Databricks

The Ocean Cleanup is a nonprofit looking to remove 90% of floating ocean plastic by 2040. They are looking to execute this plan by cleaning up the legacy plastic – the plastic already floating in the ocean – and by stopping the sources of plastic flowing into our oceans. The Databricks for Good - Oceans initiative has leveraged 4 pro-bono Databricks employees to support their efforts by applying Databricks’ versatile capabilities to the following areas:First, the Rivers Team at The Ocean Cleanup has built a probabilistic computer vision model to determine how much plastic is being released into the ocean. The model currently has issues around incremental ingestion and horizontal scaling. We have deployed two Databricks RSAs (Patrick Leahey, Michael Berk) to improve pipeline efficiency and model development practices.Second, the Central Data + IT team at The Ocean Cleanup is looking to develop a reporting platform that correlates impact with funding. Here, we deployed RSA (Brad Jarret) to develop a prototype of this pipeline to help the organization make data-driven decisions on where to allocate incremental funding.Third, The Ocean Cleanup Data + IT department is looking to streamline their tooling and architecture. Currently the organization has a piecemeal approach to infrastructure, leading to a high level of customizability but also high complexity and maintenance costs. We have deployed an SA (Bachar Wehbi) to advise on cloud architecture with a focus on increasing deployment stability.