Recycling Better with the Help of Artificial Intelligence
Despite the proliferation of recycling schemes in recent decades, human error sees to it that a large portion of these materials actually end up getting burned or buried in landfill: it’s estimated that only three quarters of what goes in the recycling is recyclable, and worldwide only 16% of rubbish ends up getting recycled . But as for many other human-created problems, AI may be coming to the rescue. Various initiatives have shown that machines are often better at sorting rubbish than humans are, and could help us to greatly reduce the waste resulting from recycling mistakes.
Applying AI to rubbish sorting is actually nothing new - basic AI started playing a role in recycling programs as early as the 1970s. A form of basic AI has been around since then in the reverse vending machines used for bottle collection return. These machines used light sensors which could detect colours: the machines would then decide whether or not to accept an item based on the colour proportions it measured.
In recent years, however, the application of machine learning has allowed systems to start going beyond making pre-programmed decisions, often exceeding human rubbish-sorting ability in the process. The large amount of non-recyclables which end up in the recycling are a serious problem for recycling plants, but robots can be trained to reverse this contamination. One company has created a robot with touch as well as light sensors. It uses both in order to detect and remove items which don’t belong in a particular stream, and can even do this when an item doesn’t look like anything it’s sorted previously.
AI has also been successful at sorting rubbish at the pre-collection stage. One company in California has trialled using a robot to sort its rubbish into recycling and other waste categories at their office. Over several months, levels of sorting contamination were cut to below 5 percent from 20 percent. Other technology forgoes the use of robots, but still makes use of artificial intelligence to guide humans on where to throw something away. One system combines a camera, which the user shows an item of trash to, with a screen which then tells them where the item belongs. If the trash is correctly sorted, a congratulatory message appears on the screen, and a QR code for a discount or other reward appears.
Although Junk2Go may be largely human-powered, we still can’t be beaten when it comes to handling that junk removal that’s too tough for regular rubbish collection. As an affordable alternative to cheap bin hire, we do the work for you, ensuring that your leftover junk gets recycled wherever possible. Call 0800 586 524 today.