Manufacturers of machinery used in the textiles-manufacturing industry confront many challenges – new materials, new products and applications, stricter tolerances for product quality and process stability, and increasingly high requirements for resource efficiency all drive the need for innovative production solutions.

At the same time, the textiles industry is a highly competitive market and any production machine must be able to demonstrate its value to justify investment – and that is especially true of machines for nonwovens. “We want to continuously improve on the value we provide to our users, in terms of machine and line performance but also in terms of usability” – says Rebekka Dilo, Head of DiloGroup’s Technical Research Center. She continues: “One area where we definitely see a growing awareness in the market is resource efficiency, especially energy and fiber efficiency”. The company is a leading manufacturer of needlefelt production lines. It serves customers in more than 80 countries worldwide and is constantly working to provide solutions that help users optimize their entire production process.
Boosting reliability and cost efficiency in the textile industry
Assisting users with artificial intelligence
What makes such optimization so challenging is that nonwovens production involves several complex processes that are sensitive to changing production conditions. Identifying correlations and reacting properly to changing production conditions can be difficult.
As many companies will see many of their experienced machine operators retire, transferring all of that know-how will soon become a challenge, “which is why we are currently exploring how we can preserve the experience that operators have as a technical solution to better support our customers,” says Dilo. The solution DiloGroup is currently evaluating for this purpose is the Siemens Cloud application Smart Machine Assistant, a self-learning application that uses machine learning capabilities to determine the optimal settings of an industrial machine in a complex environment. “Product quality of nonwovens depends on many parameters, with time lags between cause and effect and multiple interdependencies” – explains Dilo. This makes formalizing machine and process behavior very complex, and any optimization requires large amounts of data, “which is why having a self-learning algorithm was a big help” – says Dilo.