Stopping textile defects before they become wasted fabric.
Photo credit: Smartex, via SOSV
Smartex is a Porto, Portugal-based industrial technology company that equips circular-knitting machines with cameras, lighting, edge computing, and machine-learning software. Founded in 2018 by Paulo Ribeiro, Gilberto Loureiro, and António Rocha, the company developed its CORE inspection system to identify holes, dropped stitches, oil marks, yarn irregularities, and other fabric defects while knitting is still underway rather than after an entire roll has been produced.
The platform combines real-time inspection with production monitoring. When the software identifies a serious defect, its “Golden Stop” workflow can alert an operator or stop the machine so the problem can be corrected before more damaged fabric is knitted. Smartex also acquired Turkish textile-technology company Tuvis and has expanded into major knitting markets including Portugal, Turkey, Bangladesh, China, and the Americas.
Smartex announced a $24.7 million Series A in 2022, led by Lightspeed Venture Partners and Tony Fadell’s Build Collective, after earlier HAX and other funding. The company has named manufacturers and industry partners, but it does not publish audited installation counts, defect-detection accuracy by fabric type, or customer-level yield improvements. Its cumulative avoided-waste figures are company calculations and should not be confused with independently weighed landfill diversion.
Smartex intervenes at the manufacturing stage where defects can propagate through a fabric roll. Detecting a fault quickly can reduce off-quality fabric, avoid wasting yarn and machine time, and prevent the embedded water, energy, dyeing, and emissions associated with material that cannot be sold at its intended grade. Production data can also give mills a more consistent record of quality than periodic manual inspection.
Actual impact varies with machine utilization, textile construction, defect type, operator response, and what would otherwise happen to off-quality fabric. Smartex’s water, carbon, energy, and garment-equivalent figures are modeled conversions from estimated fabric avoided, not direct measurements at every mill. The system also consumes hardware and electricity, and no public comparative life-cycle assessment quantifies its net footprint or establishes that every flagged defect would have become waste.
Smartex reported in March 2024 that its systems had prevented 1,000,000 kilograms of fabric waste, which it translated into approximately 5.88 million T-shirts. By September 2025, trade reporting citing the company put the cumulative figure above 1.1 million kilograms and associated it with modeled savings of 8.1 billion liters of water and 6,739 metric tons of carbon dioxide. These are attributed company estimates, not independently audited environmental accounts.
Reporting in 2025 also cited more than 110,000 “Golden Stops.” A stop is an intervention after a detected problem; it is not equivalent to a unique defect, a verified kilogram saved, or a garment diverted from landfill. CNBC and Portugal’s trade and investment agency repeated the one-million-kilogram claim while clearly attributing it to Smartex.
The company’s demonstrated achievement is commercial deployment of machine-vision inspection and production monitoring in operating textile mills. Broader claims about eliminating textile waste, universal defect detection, or industry-wide water and carbon savings remain goals or modeled extrapolations rather than measured outcomes.
In-line defect detection
Cameras and machine-learning models inspect knitted fabric continuously, identifying defects earlier than end-of-roll manual checks.
Automatic production intervention
Golden Stop workflows can notify operators or halt a knitting machine before a localized fault propagates through more fabric.
Mill-level production visibility
Monitoring software records machine performance and quality events so mills can investigate recurring faults and compare production runs.
Avoided embedded resources
Preventing defective fabric can preserve the yarn, water, energy, labor, and machine time already invested before later finishing stages.
Locations
