Field report SW-76 / P6 · sheet /operations
Operations & RestructuringFiled from the territory
AI fabric checks could save H&M $100 million annually
Smartex.AI, backed by Tony Fadell and H&M, is rolling out AI cameras to inspect fabric three times faster than workers, with waste reduction potentially saving H&M $100 million a year.
- Grid
- SW-76 / P6
- Surveyed
- November 19, 2026
- Reading
- 4 min read
- Correspondent
- Sophie Lindqvist

Field notes
- Smartex estimates fabric-waste reduction could save H&M $100 million annually.
- Fadell invested roughly $10 million and is becoming Smartex chairman.
- Smartex Inspect is claimed to inspect fabric three times faster than a human.
- H&M reportedly handles 2.5 billion meters of fabric annually at roughly $4 per meter.
- Smartex has 65 employees and 800 older systems installed across 10 countries.
By Sophie Lindqvist
Smartex.AI estimates that cutting H&M's fabric waste could save $100 million a year, as the company rolls out AI-powered fabric inspection it claims runs three times faster than a human. The projection is substantial, but Smartex has disclosed neither a deployment timetable nor independently verified savings.
Tony Fadell, the co-designer of the iPod, iPhone and Nest doorbell, has invested roughly $10 million in Smartex and is set to become its chairman. H&M has also invested and is a launch customer, placing the company cited for the projected savings among the first users.
Smartex employs about 65 people and generates annual revenue in the low millions. It has installed 800 older monitoring systems across 10 countries; its valuation stood at $55 million four years ago, although CEO Max Easton declined to provide an updated figure.
Why fashion inspection remains manual
Behind fashion's emphasis on speed and trend, fragmented suppliers and aging warehouses still shape production. Some facilities lack even internet access, while much of the inspection work remains physical and repetitive.
Most fabric inspection depends on workers comparing swatches with cloth moving past them. They search for obvious defects, including holes, needles and incorrect yarn, while judging whether a white T-shirt matches an exact shade, from brilliant white to creamy off-white.
At Web Summit in Lisbon, Easton said: “Most people are just kind of amazed by how chaotic it is.” The disorder extends beyond the factory floor. Brands depend on independent operators across several foreign countries, leaving them without dependable shipment timing or visibility into conditions when material reaches them.
Fadell said a mill or garment facility today can look almost identical to one from 50 years ago because its processes and equipment have not been upgraded. Some sites have email, while others lack even Wi-Fi.
Fadell described fabric inspection as “very, very technical, and it’s highly prone to human error.” He said judgments can shift with the time of day or day of week, making them subjective; unlike color matching, quality inspection has no equivalent Pantone benchmark.
Two products extend the system
Smartex is rolling out two linked products as it moves beyond its installed monitoring base. Together, they connect machine-level inspection with supply-chain analysis.
- Smartex Inspect is a physical AI camera system, meaning camera-based analysis embedded in machinery, that attaches to existing textile equipment. It checks material continuously as it passes.
- Smartex IQ generates supply-chain intelligence, meaning analysis of data from Smartex's product suite, giving customers a wider view of sourcing and production.
Calibration is the difficult task. Lighting, fabric speed and small color variations affect images, so the system must calibrate against each client's standards. Smartex says its library contains 50 billion fabric images for those comparisons.
Fadell, who has backed the venture for roughly a decade, frames the ambition as AI applied to physical operations that customers can use easily. His warning is practical: a technically strong system still fails if factories find it cumbersome.
Where the $100 million comes from
Fabric represents 50% of an apparel company's cost base. Because quality varies, H&M typically orders about 5% more fabric than required, allowing for material it may reject or waste.
Easton estimates that H&M handles 2.5 billion meters of fabric globally each year. At $4 per meter, that represents about $10 billion in fabric costs; he believes better inspection could remove $100 million of annual waste.
Fadell said H&M's priorities are price, speed, quality and sustainability. The company often pre-sells garments online, yet delivery can fail when defective fabric reaches the garment house and production stops. In his view, reducing waste directly improves margins.
What does this change for H&M?
If Smartex meets the claim, H&M can order closer to actual need, reduce rejected material and protect margins. It could also prevent some online orders from failing because fabric did not pass inspection. The operating change is greater purchasing precision, not a new consumer-facing product.
The caveat is material. Smartex has not disclosed how many mills will use Inspect, how quickly the rollout will reach H&M, or what independent evidence supports the three-times-faster and $100 million claims. The $55 million valuation from four years ago, low-million revenue and 800 older systems depict an early-stage vendor.
For Fadell, the $10 million investment and chairman role show conviction. For H&M, the relevant test is operational: fewer surplus meters, fewer defective shipments and more fulfilled online orders. Until those measures appear in disclosed results, the $100 million remains a management estimate rather than a booked saving.
via fortune.com/feed/ (Source)
More from Sophie Lindqvist
Show full bio
Correspondent covering money, funding and market shifts around corporate strategy and chief-executive leadership at CEO Frontier.
483 articles
Adjacent sheets
- C.H. Robinson Credits AI With 60% Productivity Gains Since Late 2022
- Amazon Blocks Meta's Muse: The Battle to Replace the Smartphone Has Begun
- UAE Leads World in AI Use at 73.3%, but Only 29% of Firms Pay
- Companies replace blanket AI access with budgets and controls
- Brands' Next Customer Is an AI Agent—and It Ignores Brand Size