The Intelligent Factory: How AI and Digital Innovation are Tackling Textile Waste
Credit: iStock
12 August 2025
Every year, despite decades of technological advancement, the textile industry continues to generate millions of tonnes of waste annually, and the problem has only intensified over time. What makes this even more problematic is that a significant portion of this waste is never tracked, traced, or accounted for in sustainability reports.
Instead, we know that vast quantities of textile waste are offloaded each year to regions like East Africa and Southeast Asia, regions that often lack the infrastructure, financing and regulatory frameworks to manage it sustainably. Once exported, these materials largely disappear from the supply chain’s radar. Informal markets may repurpose a fraction of it, but the majority either piles up in open dumpsites or pollutes local ecosystems. It’s an invisible cost that’s borne not just by the environment, but by the communities living near these sites.
HOW CAN WE CLASSIFY TEXTILE WASTE?
While discarded clothing often dominates the conversation, a significant amount of waste is created before products ever reach store shelves. Waste occurs at every stage of the production process, from fabric cutting to garment assembly, long before consumers are involved. In fact, textile waste is often categorised into three main types, each originating at different stages of the value chain:
- Post-industrial waste is generated during the garment manufacturing process itself, once fabric has been produced and is cut, sewn, and assembled into finished products, but before they leave the factory. This category includes fabric off-cuts, misprints, defective textiles and surplus trims. Often produced within factory settings, post-industrial waste is relatively centralised but frequently overlooked despite its substantial volume and high recovery potential.
- Pre-consumer waste consists of unsold, excess, or returned garments and textiles from retail outlets or wholesale distributors. Unlike post-industrial waste, pre-consumer waste has already been finished and often retains a high degree of quality, but it remains unsold or unused due to shifts in demand, changing fashion trends, or seasonal stock clearances. This waste can include surplus inventory, sample garments, or returns that brands or retailers decide not to put back on the market.
- Post-consumer waste is clothing and textiles discarded by consumers. This is considered to be the most visible type of textile waste and includes garments that are thrown away or donated after use. This is considered the hardest category to sort, categorise, redistribute or recycle.
In this article, we will be focusing specifically on post-industrial waste: while all these types contribute to the overall impact of waste, this type of waste is a critical focus for intervention and circularity efforts. Often seen as an unavoidable by-product of production, too time-consuming or costly to manage effectively, post-industrial waste is frequently incinerated, dumped, or left to accumulate in storage, with little to no traceability.
And yet, amidst this challenge lies a major opportunity, as the manufacturing stage represents one of the most controllable and high-leverage points for reducing waste production.

Source: iStock
THE CURRENT CHALLENGE: WHY POST-INDUSTRIAL WASTE GOES UNUSED
Think of post-industrial waste as existing within what we might call the “controlled environment” of manufacturing. Unlike post-consumer waste, which relies heavily on individual behaviour and collection systems, post-industrial waste is concentrated, predictable, and generated within existing systems of production. This predictability creates unique opportunities for systematic intervention.
Currently, there are several reasons why post-industrial waste goes unused:
-
- The first challenge is perceptual: waste is often seen as not valuable “enough” to justify the effort required to manage it effectively. This creates a self-reinforcing cycle where lack of investment in waste management systems perpetuates the perception that waste has no value, when in reality, it has an incredible potential that remains unrealised.
-
- A manual process: factories rely on time-intensive manual processes to sort waste materials by fibre type, colour, and quality. This creates a bottleneck where the human effort required to make waste useful exceeds the perceived value of the materials. It’s like having a treasure chest that requires hours and hours of sorting through debris to access: the treasure might be valuable, but the effort barrier is too high to pursue it.
-
- Most factories operate without real-time tracking systems for their waste streams, creating what we might call “institutional blindness.” They can’t efficiently identify what waste materials they’re generating, when they’re producing them, or in what quantities. This invisibility prevents systematic approaches to waste management: you can’t optimise what you can’t measure.
-
- The absence of centralised systems to aggregate data represents perhaps the most significant structural barrier. Individual manufacturers may generate waste streams that are too small to attract recyclers independently, but when aggregated across multiple facilities, these streams could support viable recycling operations. Platforms like World of Waste are addressing this aggregation challenge by mapping textile waste in key manufacturing countries, helping to bridge the information gap and make these flows visible.
-
- Time pressure and lack of incentives often lead factories to dump or burn waste materials rather than invest in proper segregation and matching systems. When production schedules are tight and waste management is seen as non-essential, the easiest option becomes disposal rather than value recovery. This is particularly problematic because it treats waste management as separate from production optimisation rather than as a component of efficient, responsible manufacturing.

DIGITAL SOLUTIONS ARE CHANGING THE GAME IN THE WASTE INDUSTRY
While what we call hard tech solutions like fibre-to-fibre recycling often capture attention because of their capabilities in transforming waste into a resource, they typically deal with waste after it’s been created. Soft tech innovations like digital tools that optimise processes, reduce errors, and generate real-time insights are increasingly proving their value by intervening before waste is produced in the first place.
This is especially relevant given the mounting regulatory pressure, such as the tightening of greenwashing laws, and a general sense of stagnation in progress toward circularity. Moreover, soft tech solutions typically require significantly less capital investment than their hard tech counterparts, making them easier to scale across organisations and industries.
One significant source of textile waste at the manufacturing stage comes from poor material selection and inefficient sampling processes. Traditional fabric development involves creating countless physical samples and prototypes, many of which never make it to final production. This sampling waste is compounded by miscommunication between design teams, suppliers, and manufacturers about material specifications, leading to costly revisions and rejected batches.
Swatchbook tackles this challenge by digitising the entire material selection process. The platform creates highly accurate digital representations of fabrics, capturing visual appearance and technical properties like drape, texture, and performance characteristics. By building comprehensive digital material libraries, Swatchbook enables design teams to make informed decisions earlier in the development process, reducing the need for multiple physical samples and iterations.
Another major cause of textile waste lies in quality control during manufacturing. In conventional production, fabric defects are often detected too late, even after meters of flawed material have already been produced. These defects can result from issues like yarn breakage, tension inconsistencies, or equipment malfunctions, and typically go unnoticed until manual inspection. The result? Entire rolls of fabric are discarded, even if only partially damaged.
Smartex addresses this problem with its IoT sensors and AI/machine learning software. Installed directly onto circular knitting machines, Smartex’s technology uses advanced computer vision to detect defects in real-time: the system analyses billions of defect images, comparing each output against known patterns to trigger immediate alerts or machine stoppages when issues arise. This allows operators to act immediately, pausing production, adjusting settings, or removing flawed material, before waste accumulates. The result is a significant reduction in faulty output and material loss.
Other Articles
Fashion for Good Launches “Beyond50 Denim” to Address Hemp Integration Barriers In Global Denim Production
The Circularity Race: How Fashion for Good is Changing the Footwear Industry