To some, recycling is a waste of money–a bow to do-gooders that does little for the environment. But that’s not how it looks to Eugenio Garnica. For those on the cutting edge of robotics, recycling is just another problem about to be solved by pattern recognition and deep learning.
Garnica leads a company in Spain called Sadako Technologies that’s repurposing robots for the recycling conveyor belt. By training machines to spot valuable materials, he says it’s possible to cut the cost of sorting by more than half and thus fundamentally change the economics of the industry.
The thing is that robots actually like “3-D” jobs that are “dirty, dull, and dangerous.” They don’t mind working around the clock, and they don’t get injured when some idiot leaves hypodermic needles in the trash. Robots are perfect for recycling, and in the next few years, they’ll probably start taking over. Unlike certain industries where robots taking over might be a problem, it’ll actually be a blessing for the recycling business, not to mention the municipalities and taxpayers it relies upon.
Sadako’s SCARA robot uses compressed air suction to pluck things from the belt. It’s already installed in two plants near Barcelona with a third order on the way. Garnica says companies can make back the cost of the machine in about a year.
“We started this development because in many treatment plants some material was being thrown without being collected,” he says. “It wasn’t cost-effective to use [traditional] machines for some streams and the material has less worth than the salary you need to pay someone.”
Robots have been used in car and food plants for years. The challenge for Garnica is to apply the necessary intelligence to the recycling machine so it understands what it needs to retrieve. He does this by exposing it to hours of conveyor belt action, so it begins to learn the characteristics of each object, and so it can decide when it’s best to swoop down to grab something.
We covered another couple of robot recyclers here. Don’t count the recycling industry out just yet–automation will make the job easier than it is today.