French innovation network R3ilab explores use of AI in textile production
Last year, French research network R3ilab, a network of textile, fashion and creative industry leaders promoting intangible innovation, set up the Texia collaborative programme, aimed at identifying the opportunities for using AI tools in the textile, leather and apparel industries. Four French companies operating in these sectors have been chosen by R3ilab to join the programme, collaborating with technology partners in the field and developing solutions that can be used as a stimulus for the industry at large.
The four projects were presented on January 30 at the Parisian headquarters of Cap Digital, a European association of digital innovators, a partner in the initiative. “We received 23 applications by [textile and apparel] companies, from which we drew a shortlist of 14 based on criteria such as funding availability and whether the project was truly AI-based, and not simply tangential to it,” said Stanislas Vandier, programme coordinator with Isabelle de Bussac. “We received applications which represented a significant cross-section of the various potential uses of AI in the industry,” he added.
The first project was developed by Segard Masurel, a 173-year-old northern France specialist in wool trading and transformation, which explored AI-driven solutions to manage the volatility of markets, prices and inventory levels. “Our market is extremely volatile, and one of our challenges is how to better anticipate demand, and adapt production accordingly,” said Olivier Segard, president of Segard Masurel, a company with 70 employees and annual revenue of €125 million.
Another material, another challenge, with Lemaitre Demeestere, the last French fabric manufacturer specialising exclusively in linen. The company, with 32 employees and revenue of €4 million, is collaborating with trend forecasting expert Heuritech to develop solutions for adjusting its production activity depending on trend forecasts derived chiefly from social media monitoring. “AI can be a means of predicting risk, and also the needs of our clients,” said Olivier Ducatillion of Lemaitre Demeestere.
The third project was put forward by yarn spinner and dyer UTT Yarn. In partnership with data specialist TLG Pro, UTT Yarn is using AI to optimise its manufacturing process and the recycling of water used in the dye baths. The company, which has 130 employees and generates a revenue of €25 million, has used AI tools to build its own recycling plant. “They have a large number of sensors and data, but the process isn’t yet centralised,” said Marc Verwicht, business developer at TLG Pro.
The fourth and final project was presented by home linen specialist Garnier-Thiebaut. The company, which has 220 employees and revenue of €20 million, is relying on AI tools to identify new markets relevant to its line of business and expertise, thanks to Braincube, an industry 4.0 performance solution. “We have a comprehensive organisational structure, from yarn sourcing to logistics, and AI tools help us with knowledge management,” said Production Director Bertrand Plaze.
AI for the textile industry still “work-in-progress”
“AI wasn't exactly a speciality of R3ilab, and we needed an authority in the field like Cap Digital to make sure we were on the right track,” said Nelly Rodi, co-president of R3ilab. “We are only at the beginning of this road; it is a work-in-progress that will take two to three years to finalise. Downstream in the industry, AI tools are already utilised to identify best-selling products and aid in the creation of smaller collections that reach the market faster. But in order for [these collections] to be launched, producers need to supply the right materials, colours and creative details,” added Rodi.
This naturally raises the question of the role that AI is likely to play in textile design in the future. “Let me tell you that our collection designer isn’t necessarily enthusiastic about the programme,” said Ducatillion of Lemaitre Demeestere. “But the role of AI is not to replace creativity. It’s used to reduce risk, and to turn our attention to things we hadn’t thought about,” he added.
Similarly to fashion buyers, striving to reduce purchasing and inventory volumes, textile manufacturers are focusing on AI tools to optimise their operations. Above all, the analysis of market data and of the producers’ internal data raises questions on how to utilise AI to forecast the prices and volumes of the raw materials producers need. A challenge which, according to the industry professionals present, is too big for individual companies.
“It's something that can be taken on at the level of the industry as a whole, maybe using blockchain technology to map how manufacturing is evolving and when is the best time to source linen,” said Ducatillion. According to Segard, of Segard Masurel, AI gives a chance to better identify the factors that can impact business: “In wool production, we have to deal with the vagaries of the weather, as well as with the elements faced by other sectors. At our level, even the price of the milk used to feed sheep has an impact on [wool] volumes and prices, hence on our ability to produce. It's an approach worth pursuing.”
The Texia projects were selected by a panel of experts and industry professionals, chaired by Christine Balagué, Good in Tech professor at the Institut Mines-Telecom Business School in France, and also including Nelly Rodi (of R3ilab, CCI Paris IDF), Françoise Colaitis (of Cap Digital), Yohann Petiot (of Alliance du Commerce), Yann Rivoallan (of The Other Store), François Gonnot (of Lab Innovation Lectra), and project coordinators Isabelle de Bussac and Stanislas Vandier.
Copyright © 2022 FashionNetwork.com All rights reserved.