Marine Pollution Project

AI-based solution to marine pollution

What inspired this project?

Marine pollution can severely damage marine life and their ecosystem. Millions of marine animals die from marine debris such as plastic each year. Microplastics from marine pollution can even be ingested by humans, causing damage and inflammation. Marine pollution can also disrupt ocean-related industries.

Our Method

Our app is an application built with the Unity game engine. Using a backend flask server and open data from Debris Tracker, the app can predict the location, type, and quantity of marine debris, and display them on a world map. The AI is trained from a large pool of existing publicly accessible data, which allows it to accurately and precisely predict marine debris.

Sufficient data

Our app requires an abundant amount of marine debris data, since both the app itself and the AI require a large amount of data. The data used would have to be accurate, precise, and reliable to prevent miscalculations.

Usability

Users on all platforms should be able to easily access our app. A versatile engine like Unity allows the app to be portable and run on different platforms. A simple and easy to understand UI would also be needed to make the app easy to use.

Artificial Intelligence

The AI used for this project needs to be accurate as well as consistence to provide reliable debris predictions. Thus, a reliable database of marine debris would have to be sourced to provide precise and consistence data.

Who are we

How it works

Fast

The app constantly displays the amount of marine debris. Users can filter different types of debris and hide volunteer locations using the Toggle Volunteers and Toggle Debris tools. Clicking on a debris or volunteer marker will display information about it, such as location, description, and type.

Smart

Upon, entering the app, users are presented with a worldwide map of marine debris and volunteer locations spawned from existing data. Users can also choose to place down their own markers to be predicted by the AI.