Eurecat — Photonic technology and artificial intelligence to predict defects in textile manufacturing

Photo - Eurecat

The Eurecat technology center is showing at ITMA (International Textile Machinery) 2023, the leading fair of the textile industry that is held in Milan until June 14, a solution based on photonic and Internet of Things systems, combined with artificial intelligence algorithms, which make it possible to predict potential manufacturing defects, before fabric manufacturing, and inform the person responsible for production in order to evaluate them and make decisions about them.

The innovation, which has been developed specifically for large-diameter circular knitting machines in collaboration with textile machine manufacturer Canmartex, “contributes to boosting the sustainability of the industry, increasing production and reducing costs, as well as to improve the innovative positioning of companies”, explains the director of Eurecat’s Industrial Technologies Area, Xavier Plantà.

Specifically, the technological solution is based on real-time analysis of the state of degradation of the most critical components in the manufacturing process, such as needles and plates.

It is estimated that 92 million tonnes of waste are generated annually from the textile industry, 25 percent of which is produced during the manufacture of the fabric. In this context, the aim is to perform predictive control using digital technologies to “avoid typical quality problems or defects such as holes, scratches and stains, among others, which are generally detected after the manufacturing process by expert people or by automated systems based on artificial vision, once the defects have already occurred”, adds Xavier Plantà.

According to the director of Eurecat’s Industrial Technologies Area, “anticipating before the defect occurs and, above all, acting and correcting the causes that cause them is the focus and strategy of technology developed, along the lines of the slogan chosen for this ‘Better to predict than to correct'”.


Please enter your comment!
Please enter your name here