Abstract

Plastic pollution is a big environmental challenge facing us today. We use a huge amount of plastic in our daily lives but recycle very little. In fact, the average person living in North America consumes around 230lbs (104kg) of plastic each year! When a problem is this big, it can sometimes feel overwhelming. But if everybody makes small changes, it soon adds up to something big!

Recycling old plastic helps to reduce the amount of new plastic we make. However, it is difficult to identify and separate all the different types of plastic by simply looking at them. We wanted to find a solution to this. Our study shows that computers with cameras can learn to recognize 12 types of plastic. They can then use this knowledge to identify plastics as they move along a conveyor belt. Our findings could revolutionize plastic recycling and bring us one step closer to solving the global plastic challenge.