Fashion’s waste problem begins on the factory floor. Smartex uses AI and cameras to spot fabric defects during production, helping factories reduce waste.
Fashion has a waste problem. Every second, about a truckload of clothes is thrown out—buried or burned—according to the Ellen MacArthur Foundation. Behind the scenes, much of that waste starts earlier in the production process. Fabric is lost before it ever becomes clothing.
That’s the problem Smartex wants to solve. The company builds camera-based systems that use artificial intelligence to detect defects in textiles as they are produced. By catching problems early, factories waste less material. Over the past three years, Smartex says its tech has helped prevent 1 million kilograms of fabric from being thrown away.
The company claims its system improves production efficiency by around 0.37% per kilogram of finished fabric. That may sound small, but it adds up. In 2024, fashion giant Inditex (which owns Zara) used more than 678,000 tonnes of raw materials. A fraction of a per cent saved at that scale means more garments made, less waste created.
Part of the challenge is that garment production is still highly manual. Supply chains are long and fragmented. Cotton may be grown in one country, dyed in another, sewn in a third. That complexity makes standardising or digitising the process difficult. At the same time, the fashion industry moves quickly and reacts to shifting trends.
In an interview with CNBC, Smartex co-founder and CEO Gilberto Loureiro said this lack of technology in factories is a missed opportunity. “If this is the largest industry that is still untouched by [the] internet and is one of the largest pollutants in the world, and nobody is working on this in terms of technology [then] there is a massive gap here,” he said.
The environmental toll is serious. The EU says 20% of global water pollution comes from dyeing and finishing fabrics. Reducing waste and improving quality control at the source could ease pressure on both resources and margins.
Smartex has caught the attention of investors. In 2022, it raised $24.7 million in a round led by Tony Fadell, the inventor of the iPod, through Lightspeed Venture Partners. H&M Group also backed the company that year. Smartex has raised over $40 million in total. Loureiro told CNBC he sees strong potential in the space, even if it’s complex. “There is a huge value to capture. So that’s a big risk, being reward,” he said.
Amazon has also invested through its AWS Compute for Climate Fellowship, a program supporting startups focused on climate, food security, and conservation. Smartex uses AWS infrastructure to train its machine learning models. These models must adapt to variations in fabric and production environments across countries and factories.
Loureiro, speaking to CNBC, said he spends much of his time visiting factories in Bangladesh, Vietnam, and other manufacturing hubs. Factory owners are focused on one thing—payback. “If in 30 seconds he’s not convinced about the ROI [return on investment] payback, in less than one year, for example, you are out of the game,” he said.
Most customers spend several hundred thousand dollars on the Smartex system. Loureiro said the tech must prove its value quickly through savings in materials, energy, and time. According to the Apparel Impact Institute, the average return period is nine to 18 months.
Long term, Smartex has broader goals. It wants to become the digital layer that connects all parts of the fashion supply chain. Loureiro calls it an “operating system” for factories—software that helps brands track where items come from, how they’re made, and what resources were used. “These are basic questions that are very difficult or impossible to answer by most fashion brands,” he told CNBC.
Fadell has likened the idea to how Apple’s devices work together through shared software. Loureiro explained, “It’s not about the computer, the Mac or the iPhone or the AirPods, it’s about what they can do all together.”
For now, Smartex is focused on reducing waste and helping factories run more efficiently. It’s a technical fix to a longstanding problem—one that the fashion industry is only just beginning to confront.
(Photo by Unsplash)
