AI to identify recyclable waste

Why this application ? How does it work ? How to make it real ?

Why this application ?

In 2017 the former French Prime Minister Edouard Philippe has announced his willingness to recycle 100% of plastic waste. Even if this goal seems utopian, it is possible to increase the amount of waste recycled considerably, knowing that it was only 26 per cent in 2016. In order to do so, consumers have a significant role to play. This starts by knowing which plastic waste can be recycled and which can't. This is why we launch this application, which will be able to identify recyclable plastic waste.


How does it work ?

This application will use artificial intelligence, and more precisely supervised machine learning. Supervised learning is an approach to creating artificial intelligence (AI), where the program is given labeled input data and the expected output results.The system is fed with massive amounts of data during its training phase, which instruct the system what output should be obtained from each specific input value. The trained model is then presented with test data to verify the result of the training and measure the accuracy. In concrete terms, we are going to provide a large amount of picture containing plastic waste and precise each time if it is recyclable or not (the input). Once the system is trained enough, it will be able to recognize by himself if a plastic waste can be recyled or not (the output).

Recyclable
Non-recyclable

All the user will then have to do is take a photo of their waste to find out wether or not to recycle it.

How to make it real ?

You can find a lot of resource on the web which can help you to create this application. The most useful should be the imageClassifier() method from the ml5 library. To learn how to use it step by step, you can watch this video. And here is to know which plastic can be recycled :