AI’s Role in Material Innovation – Part 1 The - Première Vision Paris - Denim Première Vision - Première Vision New York
What is innovation? While this subjective question can be answered in many ways within the textile and fashion industry, one must objectively agree on one point: the indubitable potential for the diversification of materials. Materials lie at the heart of the fashion industry. They account for 92% of the industry’s total emissions through their extraction, processing, and production1, and around 30% of the cost of goods sold (COGS)2. Therefore, in parallel to the primary imperative of reducing greenhouse gas (GHG) emissions, resource consumption, and the pressure on biodiversity—while freeing the industry from fossil fuels—the drive towards circularity and the research and development of so-called “Next-Gen” materials represent one of the many complementary actions aimed at creating a less impactful fashion sector.
Especially in the face of climate change, resource scarcity, geopolitical disruptions, and forthcoming regulations, the material transition is more crucial than ever. The urgency to effectively implement the material transition is at its peak, given that most of these materials are still in the pilot phase. However, many brands lack the preparedness required for this move and would need guidance for driving adoption to unlock the benefits of these new materials.
Together with Boston Consulting Group (BCG), Fashion For Good released a new report, Scaling Next-Gen Materials: an Executive Guide 3 based on industry best practices and the successes of early movers. This report provides brands with a structured framework to act on key levers and outlines a strategic pathway to address major challenges while driving measurable results.
The research reveals that next-generation materials could represent 8% of the total fibre market by 2030—equivalent to approximately 12.5 million tons.
“The challenge lies in bridging the gap between innovation and accessibility. Scaling next-gen materials requires collaboration across the value chain, investment in infrastructure, and a commitment to reimagining how we design and produce at every level. […] Novel and innovative fibres and materials with desired improved environmental and/or social outcomes when compared with conventional options; are currently in early stages of commercialisation or development; and require further technological advancement and cost optimisation for widespread adoption. This next generation of fibres and materials promises to advance solutions for the sustainability and circularity challenges faced by the industry. The vision is for these materials to transition from ‘next generation’ to ‘preferred existing’ status.”
– Fashion For Good, Scaling Next-Gen Materials: an Executive Guide
Next-gen materials
In this first chapter of a two-fold series of articles on material innovation, we focus on next-generation fibres that replace conventional man-made fibres. We examine this new generation of fibres that are not dependent on virgin resources, instead using waste from food, agriculture, or textiles as their feedstock. One might begin with Cupro, one of the oldest examples, which utilises cotton linters—a waste by-product of the cotton industry—to create an artificial cellulosic fibre, comercialised under the names “Cupro™” or “Bemberg™”.
AI-Driven molecular and material design
Viewed through the lens of technological innovation, artificial intelligence (AI) could play a valuable role at multiple stages of material development, particularly as processes scale and refine.
AI and machine learning algorithms can sift through vast databases of material properties to predict which combinations of fibres will yield optimal performance. They can also accelerate the discovery and optimisation of new biomaterials by predicting molecular interactions and properties. For instance, AI can be used to model protein structures, simulate their behaviour in textile applications, and fine-tune bio-based polymers for specific performance characteristics, durability or specific degradation profiles.
One illustrative example is Epoch Biodesign, which provides an end-of-life solution for complex pre- and post-consumer polyamide (nylon) waste—from elastane-blended sportswear to high-performance, multilayer laminate waterproofs—by producing the chemical building blocks necessary to manufacture new recycled polyamide material. They employ generative AI models, trained on the “language of biology”, to design novel enzymes capable of unlocking superior depolymerisation processes. This involves generating thousands of candidate enzyme designs using their protein AI models, synthesising these candidates in the laboratory, and testing them in multiplex using advanced techniques to determine which variants perform better or worse. The resulting data is then fed back into the AI model, which uses it to predict the next set of improved variants, repeating the cycle until an enzyme with the desired performance and characteristics is developed.
Another pioneering company is LanzaTech, which incorporates AI in its process of producing monoethylene glycol (MEG) from carbon emissions. By utilising synthetic biology and AI tools, they have discovered multiple novel pathways to directly convert carbon emissions into MEG through fermentation—eliminating the need for an ethanol intermediate. This innovation facilitates the production of polyethylene terephthalate (PET) resin, fibres, and bottles from captured carbon, offering alternatives to traditional fossil-based methods.
“While there is no organism in nature known to produce MEG, through this proof-of-concept stage, LanzaTech has used synthetic biology and AI tools to discover multiple novel pathways to make MEG directly from carbon emissions. By combining and prototyping various sets of enzymes identified from different sources in novel ways, LanzaTech has successfully reprogrammed its ethanol-producing bacteria to fix and channel carbon into MEG.“
– LanzaTech
Process optimisation, waste reduction, and end-of-life predictions
AI can assist in predicting and detecting material defects before large-scale production commences. In the development of next-gen man-made fibres, machine learning algorithms can analyse real-time production data, ensuring quality control and consistency by detecting molecular inconsistencies that might affect material properties. AI can also streamline production by adjusting parameters to maximise material yield while minimising energy or resource consumption. For example, one could envision that an Italian company such as Orange Fibre might employ AI-driven modelling to refine the extraction of cellulose from orange peels, reducing waste while maximising fibre yield.
In supply chain and production forecasting, AI-driven modelling can help companies source raw materials sustainably and forecast production needs accurately. AI can also be used to track material lifecycles and improve circularity models—for instance, by employing predictive modelling to determine how recycled content integrates into new materials or by simulating biodegradation patterns to ensure that a material degrades efficiently under various environmental conditions. Similarly, recycling companies could benefit significantly from these technologies to calculate risks over the whole supply and logistical chain, verify outcomes, and test alternatives and strategies before launching.
Thus, AI could be helpful in moving from R&D to full-scale production, refining biomaterial performance, minimising environmental impact, and integrating these materials into existing supply chains. While we will dive into the use of AI for the development of bioengineered materials in the next article, it is important to note that:
“The strategic adoption of next-gen materials could lead to an approximate 4% reduction in COGS over five years, compared to inaction. This demonstrates the importance of transitioning to these materials for brands aiming to maintain their competitive edge.”
– Fashion For Good, Scaling Next-Gen Materials: an Executive Guide
References:
1 World Resources Institute and Apparel Impact Institute, Roadmap to Net Zero: Delivering Science-Based Targets in the Apparel Sector, 2021
2 BCG Analysis
3 https://www.fashionforgood.com/report/scaling-next-gen-materials-in-fashion-an-executive-guide/