High-Level Project Summary
Marine plastic pollution is at an all-time high. A person could be ingesting between 39,000 to 74,000 microplastic particles a year. More than 800 species of marine life are heavily impacted.mermaidAI aims to reduce marine debris by remotely detecting, quantifying and understanding its causes. It supports 2 approaches: The above water remote sensing that uses open-source live data and AI to record volumes of pollutants, as well as continually update its dataset. The second approach works underwater; measuring water parameters and photographing debris. AI is then used to analyse and record individual pollutants.This data is made available by app so the public is aware and willing to help.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
mermaidAI supports 2 approaches:
- Analysing data from remote-sensing satellites that work above water to detect pollution
- Measuring water parameters at the surface, as well as imaging objects it recognises as debris underwater
Approach 1: Remote sensors capture variations in inherent optical properties (IOP) such as absorption and scattering. Reflection, absorption, and transmittance of electromagnetic radiation are highly dependent on the concentrations, types, and presence of substances in water. Hence, ocean color represents data that can be used to estimate the concentrations of water constituents.
Using live data and AI, mermaidAI detects and quantifies marine pollutants:
Approach 2: Open-source datasets containing various water quality parameters are used, such as temperature, salinity, ion count etc. Using AI, we can quantify pollution. High-power camera is deployed underwater. Using an image dataset of plastic pollutants, mermaidAI will record individual pieces of pollutants.
All data collected will be presented in a readable way for the public to access. They can also join clean-up efforts, and learn about impact of debris on marine life.
The ML Model was developed in Google Colab using Python. We hope to see this tool used around the world, and witness significant improvement in quality of water in our oceans.
Provided is an early draft of the App for public:
Space Agency Data
The Following provided datasets of water quality parameters:
EOSDIS Ocean: https://earthdata.nasa.gov/learn/discipline/ocean
Global Drifter Program: https://www.aoml.noaa.gov/phod/gdp/
Used for live remote-sensing data:
Global Forest Watch: https://www.globalforestwatch.org/
The following provided dataset for remote-sensing images:
CSDA Program: https://earthdata.nasa.gov/esds/csdap
AIRSAR Mission: https://airsar.jpl.nasa.gov/
Hackathon Journey
Our journey has taken us through the oceans and made us realise how deeply marine flora and fauna are affected by marine pollution. We decided it is high time we take action.
This challenge inspired us to think of new approaches and methods to solve the issue of plastic debris faced by our oceans, as well as dive deeper into the workings of AI. A setback arose with the topic of Remote-Sensing, as it was completely new to us. Fortunately, with the resources provided, we were able to learn and progress further than our expectations.
Overall, it was a very positive experience that provided us with knowledge and the drive to do more for our planet.
We're thankful for this opportunity to participate at a global level, and communicate and express our ideas. We've learnt a lot these past days, and will continue to learn!
References
https://www.iucn.org/resources/issues-briefs/marine-plastics
https://www.recode.net/ad/18027288/ai-sustainability-environment
https://www.nature.com/articles/s41893-021-00726-2#Sec7
https://www.epa.gov/trash-free-waters/toxicological-threatsplastic
https://www.libelium.com/libeliumworld/smart-water-sensors-to-monitor-water-quality-in-rivers-lakes-and-the-sea/
https://www.intechopen.com/chapters/64603
Google Colab
Tags
#marinelife #AI #remote-sensing
Global Judging
This project has been submitted for consideration during the Judging process.