Difficult, increasingly difficult, to find a niche in which artificial intelligence (AI) has not emerged in recent years. We put it in command of ships and planes, use it to create pictures and chronicles, explore the universe or entrust it to the care of our elders. And that’s among a long et cetera that includes defending in trials, stirring up the waters of political debate, or governing populations. So things… Why not take advantage of it in a crucial task, in which we have a lot at stake and that, despite the time we’ve been practicing it and the efforts to expand it, we’re still far from mastering it: recycling ?

The question is interesting and some people are already answering it.

Some data to start. We have been promoting it for some years now with campaigns, education and even prizes, but the truth is that recycling still resists us. In 2021, the urban waste rate stood at 35% in Spain, which leaves us far from the 50% target set by the European Commission and even the European average (48%). The data is based on a study by the Cotec Foundation for Innovation, but they are not the only ones who put their finger on the wound.

A year ago, Conama insisted that Spain needs to “put its batteries on” in managing its waste and reducing the percentage of garbage that ends up in landfills.

A challenge on a global scale. Not all are bad indicators, of course. Spain has improved its plastic waste management, placing —according to data from the EPRO 2020 report— among the top positions in the European ranking with an average of 13 kilograms per inhabitant. With regard to paper and cardboard waste, at least in 2019 the recycling level in our country was over 72.9%.

Nor is the challenge exclusive to Spain or its neighbors in the European Union: it is estimated that the United States alone generate 42 million tons of plastic waste per year, more than the sum of the 27 countries of the European bloc. Regardless of each nation’s data, the OECD warns that rising plastic pollution is a global challenge: less than 10% of waste ends up being recycled.

AI, a new ally. The challenge is great and some have already decided to use AI to face it. As? With robots and systems that allow you to separate waste with a higher level of efficiency than humans: faster, more safely and collecting valuable data. As Axios details, its aim is to address some of the challenges faced by the recycling industry, such as the increase in plastic waste that is difficult to manage or the mistakes that we citizens still make by throwing garbage in the wrong container.

“It gives us data to make better decisions”, explains Matt Flechter, from the Michigan Department of the Environment, about automation and the use of special software, with AI, which allows, for example, to know how contaminated the garbage is in real time. The use of technology is not exclusive to the final stage of management: recently and as part of its SmartWaste initiative, Ecoembes implemented a system in La Rioja that included sensors in the collection trucks, GPS systems and state-of-the-art technology in the plants themselves. of selection.

Machinex Samurai Artificial Intelligence

How can AI help us? Basically offering us greater efficiency. Thanks to the use of artificial vision and training, robots can help with waste sorting. “It’s like a vacuum cleaner. It knows what to pick up as things go down the conveyor belt,” Flechter tells Axios. Thanks to the sum of AI and artificial vision, for example, Picvisa allows “automate and optimize recycling plants” and “maximize performance”.

The company guarantees that its Eco Pack system, designed for the classification and separation of plastic, cardboard, metal, wood and textile waste, among others, reaches efficiency levels of up to 95%, with a “purity” of the selected material of 95 % and capacity of 12 t/h (tons per hour). The percentages are even higher on their glass device, the Eco Glass.

A million dollar business. The possibilities offered by artificial intelligence and vision, deep learning Big Data, among other technologies, has already captured the interest of other companies. And with eloquent results. At least in financing. A year ago, Glacier, a robotics company applied to recycling, announced that it had raised $4.5 million during its seed funding round.

“Today 50% of recyclable materials in the United States end up in landfills. By increasing recycling, we reduce pollution and conserve natural resources”, highlights the company, launched in 2019 and which guarantees to be able to make 45 selections per minute and classify more than 30 materials.

We've known for years that our recycling system is broken.  Looks like we're finally gonna make it right

Add and follow companies. Machinex is another company in the sector and guarantees up to 70 selections per minute, almost twice the —guarantee— of a human operator with an efficiency in the recognition of waste that reaches 95%. AMP Robotics claims even greater capacity, allowing it to sort over 80 items per minute and 150 with tandem units. A few months ago, it announced that it had raised about $91 million in corporate capital.

cover image: Vivianne Lemay (Unsplash) It is machinex

In Xataka: I was always curious about what they did with the yellow bins: so I followed one