Project Description:

The aim of the project is to develop pipelines and methodologies for mapping the extent and density of seagrass meadows in close coordination with the current Plastic-i team. The project will exploit ground truth data and satellite imagery from around the UK with the possibility of looking at seagrass meadows in other geographies.

The project will involve developing machine learning algorithms and a prototype map, which will subsequently be added to Plastic-i’s production workflows. The work will use Python as a primary programming language, with subpackages numpy, pandas, matplotlib, pytorch, xarray and xgboost being the most utilised frameworks. The applicant will develop plots that show the likelihood of seagrass being present in a region and develop machine learning algorithms to create these images.

They will have the opportunity to present their findings to the rest of the team. This is an area of active development for Plastic-i, so the applicant will be able to develop prototypes that will become production-ready systems in Plastic-i’s future technologies.

The project will take place in Summer 2024, over a period of two months.

Applicant Specification:

The applicant will be a driven individual who is able to work as part of a tram but also take the initiative when appropriate. Plastic-i believes in empowering individuals to create products they are proud of in areas they are passionate about, so working with a degree of autonomy is important.

Applicants should be third/fourth year undergraduates studying Physics, Computer Science, Geography, Mathematics, Astronomy, or any related technical field with a component related to programming.

Minimum requirements:

• Experience using Python (knowledge of numpy, pandas, matplotlib, xarray, jupyter)

• Experience using machine learning algorithms, particularly in Python (scikit-learn, tensorflow, pytorch, xgboost)

Desirable requirements:

• Experience handling satellite data

• Experience with geographic information systems (GIS), such as QGIS, ArcGIS, shapefile usage.

• Experience with code version control (git, GitHub, GitLab, etc.)

About Plastic-i

Plastic-i is an award winning startup building the world’s first platform dedicated to tracking ocean health in near real time. With insights and products being

generated to monitor marine plastic pollution, blue carbon projects, biodiversity and provide data to sustainable finance indices, Plastic-i covers a broad range of

maritime sectors. Leveraging satellite data and AI algorithms, Plastic-i is set up to aid a wide range of actors, including government agencies, NGOs, marine

applications companies, blue carbon project developers and clean-up operators.

Plastic-i was founded in October 2021 by Dr James Doherty and Dr Donal Hill after winning Hack the Planet, an ocean sustainability-focused competition run by

the Commonwealth Secretariat and Satellite Applications Catapult. The company has an impact driven mission to improve the state of Earth’s oceans and help

contribute to a sustainable future for all to enjoy.