JUST GIRLZ

1

İrem Koç

Yasemin Koç

Elif Koç

​

​

LEVERAGING AI/ML FOR PLASTIC MARINE DEBRIS

2

Water pollution is a serious situation for environment, marine life and for the world economy.

Artificial Intelligence/Machine Learning techniques could be used to monitor, detect, and quantify plastic pollution and increase our understanding for this purpose.

​

​

​

Each year 8M tons of plastic are leaking into marine ecosystems

3

90 percent of plastic that pollutes our oceans comes from 10 rivers, 6 of which are in China

SOLUTION

4

https://debristracker.org/data

There are Citizen Scientists’ efforts such as debristracker.org by Morgan Stanley in order to clean the debris to solve the problem.

​

​

They clean roughly 210 tons of plastic debris which is 0.003% of total debris each year.

5

Using NOAA’s (National Oceanic and Atmospheric Administration) datasets from 1998 to 2021 we tracked debris drifts from China shores into the ocean and nearby countries.

We also calculated 25% decreased debris effect and plotted on the map.

SOLUTION

6

We completed this proof of concept study. It could be extended for every countries’ shores and drifts of debris in order to show international effects of pollution.

​

New international laws could be arranged between governments based on the severity of the effects.

​

Citizen Scientists’ efforts could be sponsored by these governments and other international/local organizations.

​

​

​

​

SOLUTION

7

We used buoy datasets from NOAA’s official web-site. Buoy movements also show the debris movements in the ocean.

We used google maps to find the coordinates of China’s seashores. We filtered buoys which located in this area and track them.

We used AWS SageMaker Studio and Python to clean and visualize the data. We used Matplotlip, Pandas and Geopandas libraries for the visualization.

We randomly filtered out tracked buoy data by 25% and visualized it.

​

​

TECHNICAL DETAILS