Introducing TrashBot, an autonomous system that can make recycling much more efficient and affordable
Recycling: one of the most important tasks facing mankind. It conserves energy – because we can reuse old materials instead of having to manufacture new ones – therefore curbing greenhouse gas emissions and slowing the rate of climate change. It reduces the amount of waste in landfills, which has already become untenable. And recycling also cuts down the amount of plastic in our oceans (which couldn’t be more important, given the prediction that, at the rate we’re going, there will be more plastic than fish in the oceans by 2050).
However, right now, only 20% of what we put in our recycling bins actually gets recycled. One reason for this is that humans get confused about what can and can’t be recycled. Enter TrashBot – CleanRobotics system for sorting through recycling by using artificial intelligence to identify items and categorize them accurately. The idea is that, if TrashBot can be rolled out on a wide scale, it can make a big impact on preventing contamination in recycling, and raising that figure from 20% to much higher. In reducing the amount of recycling wasted in commercial recycling facilities, TrashBot can make recycling more economically efficient, too.
Below, the Colorado-based team behind Clean Robotics explains more.
Where is your team based, how big is the team, and who is involved?
We're based mainly out of Pittsburgh and Boulder, CO. We have a pretty small core team of five people with another 12 or so part-timers.
What's the global issue you're tackling?
Waste and Recycling. Mainly the lack of transparency, the low amount of recyclable materials recovered, and inadequate education efforts.
Walk us through your AI technology, top line.
We recognize items that people are throwing away at the bin level, then correctly categorize and sort it to make sure it ends up in the correct waste, compost, or recycling stream. While this may seem simple, most municipalities have wildly different rules as to what is and isn't recyclable, so recognizing whether a bottle has a cap on it, the bottle's color, and the shape of the bottle all matter when making sorting decisions.
What made you want to enter the $5M IBM Watson AI XPRIZE?
A friend of ours, Kenny Chen, made us enter. I believe he was also a big part in getting Marinus Analytics, another finalist, to participate too.
What's been your team's biggest challenge so far?
Merging AI, electronics, and hardware into a form that works well and is affordable. Also, figuring out what types of features we need to detect in order to achieve the highest rates of recycling.
Why is what your team is doing important now, and how do you see it scaling up in the future?
Because people suck at recycling. People need to be better educated on what is and isn't recyclable and on recycling and waste policy, which massively impacts the type of commodities we use as consumers. There are currently no consistent, data-driven, and useful methods of presently doing that; we're trying to be one of the first to educate and sort better where people most interact with waste. We're going to scale up by just replacing trash cans in public places with TrashBot. Our goal is to have 10,000 units deployed by 2023.
The waste and recycling industry has been a mess for decades, this all came to a head in 2018 when China decided not to take any more recyclables. This put the recycling industries around the world into a state of chaos since most systems were designed for speed and quantity, not quality. Now everyone is looking for solutions to improve the quality, whether through sorting or education. We are an excellent solution for many places, specifically high traffic public settings.
Has the $5M IBM Watson AI XPRIZE competition changed you?
Yes! It made us look deeply into the impact we are having on the environment as well as socially. It has made us much more cognizant of trying to shape our messaging around the good we are trying to do rather than just having a cool robot.
Outside of your work in revolutionizing recycling, what's an area of AI that's exciting you right now?
The medical applications of AI seem numerous and very beneficial. They have some kinks to work out around ethics and bias, but I have high hopes considering the people involved want to do the most good possible.