To a recycling plant, tons of waste arrive, often well-classified, but many other times not. Distinguishing which ones have been mistakenly thrown into the yellow or green container is one of the main tasks of the technology company Picvisa, with clients such as Ecoembes, Ecovidrio, Ferrovial, FCC, Urbaser, Efacec, Bianna, Leblan, Valoriza, and Verallia. To do this, the Barcelona-based company uses artificial vision, a field of artificial intelligence (AI) that allows extracting information from images or videos and making decisions based on that data. In a recycling plant, with the belts running non-stop, this technology allows digesting a glass bottle and rejecting a can in a matter of seconds, blowing it or making it fall to remove it from the circuit when it doesn't belong there.
"It works like the human eye," explains Picvisa's CEO, Joan Manel Casamitjana. Through cameras, the company analyzes the composition of the waste passing through the belt, sending it a light that, depending on whether it passes through the material or bounces off it, determines whether it should be there or not. The company applies artificial vision, enriching it with deep learning, to separate different types of waste, mainly plastic and glass, but it detects that textiles are already starting to gain weight. "This 2024 we will focus more on textiles. It will be one of the new recovery models, considering that today it is a material that goes to the landfill and is burned," says Casamitjana.
This new model will not only involve reusing clothes, as has long been facilitated in second-hand stores, but also recycling the material of garments, separating it according to its composition, shredding it, and re-manufacturing. In fact, more and more big brands are detailing on their labels that pieces are made with recycled cotton and polyester. There are even backpacks made from plastic bottles.
All these changes in the world of recycling are piloted by Picvisa from DFactory Barcelona, a new economy and industry 4.0 center promoted by the Consorci de la Zona Franca de Barcelona (CZFB). With a 20-year trajectory, the company has been part of the Calaf construction group since 2012, with offices in Sant Joan Despí. "It was a more traditional environment and the rumor started that it wasn't the most suitable place. When they started building DFactory, we were offered the possibility of moving here and, even before the works were finished, we had already moved in," recalls Casamitjana.
With a staff of 35 employees, Picvisa's CEO values that the new location allows them to be close to robotics, data analysis, and 3D printing companies, "a much more natural environment" for a technology company like theirs. "It facilitates collaboration and learning in new technologies," he emphasizes, "it's a hotbed of innovation and we are closer to the industry."
From software design to machinery maintenance
Picvisa's strong point, recently accredited with B Corp certification, is the development of all the necessary software programs to apply its technology in any plant, adapting to each machine and the different waste recovery processes. Being integrated into the Calaf group also allows them to cover the design and manufacture of the machines that execute their technology, as well as their commissioning and maintenance.
This flexibility, both to adapt to different plants and to cover the entire process, allows them to reach many countries, beyond Spain. The company has a strong presence in Europe, operating in France, Italy, Portugal, Malta, and Denmark, but also in Latin America, in countries such as Mexico, Brazil, and Guatemala.
Since 2018, the Barcelona-based technology company has registered profits, after undergoing a process of redefining its commercial strategy with the entry into Calaf. In 2002, it registered a turnover of 7 million euros, a figure it expects to exceed in 2023, surpassing 8 million euros.