Automated fabric inspection · Since 1971
Real-time defect detection for plain and patterned fabric, at full production speed, with AI classification and 3D defect grading.
The WebSpector system
WebSpector is Shelton Vision's automated inspection platform. It inspects every fabric type at full line speed: plain single-colour, complex patterned, and technical web materials.
Line-scan cameras, image-processing software and AI classification detect, classify and grade defects on the move, with no false alarms. Run it standalone batch-to-batch, on-loom, or integrate it directly into production processes such as printing, coating, embossing, continuous dyeing and finishing lines.
You get consistent quality, fewer claims, higher yield and verifiable performance, with reliable quality data across your supply chain.
Explore the technology →
End-to-end quality assurance
Defects get more expensive the further they travel. A fault that slips through early collects the cost, and the embodied carbon, of every process that follows it. WebSpector inspects at each value-adding stage, so defective fabric is caught before you pay to add more to it. Your overall defect rate falls, with less waste and rework.
Verify incoming greige and raw fabric is fit to process, before you commit energy, dye, chemicals and labour to material that was never going to pass.
Inspect between value-adding steps. Catch faults the moment they appear, so long-running or repeating defects are minimised before moving downstream to the next value-adding process, and your overall defect rate falls at source. Installed wherever value is added:
A final digital defect map, auto-classified and graded to your standard, drives a yield-maximising optimised cutting plan and precise re-roll control, delivering maximum fabric utilisation (usable, sellable metres) and minimum waste, from consistently reliable, verifiable quality data.
Inside the final gate
At the final gate, every detected defect flows automatically from inspection through to a yield-optimised cutting plan.
Accurate inspection across all fabric types, at full line speed.
Instant digital defect map, auto-classified and graded for commercial review.
Maximise first-quality yield, an optimised cut plan calculated for any work order, against your rules.
Precise re-roll and cutting-machine control turns the optimised cut plan into reality.
Our unique capability
Repeating patterns, jacquards, stripes and checks defeat most machine vision systems. The design itself reads as "noise." Shelton Vision's technology separates the pattern from the defect, so the system catches genuine faults without false alarms.
Why this is defensible
WebSpector is a software-and-data platform that compounds in value with every metre inspected. That is what makes the technology, and the business behind it, hard to replicate.
Two decades of labelled, real-world defect data train models that a new entrant cannot buy.
Commercial terms based on per-metre operational charging, incorporating service & support, spares, upgrades and remote diagnostics, built for long-term partnerships.
Protected pattern-separation methods solve the hardest problem in fabric inspection.
Classification sharpens as more fabric is processed.
The platform
Machine learning, algorithmic defect detection, imaging science, deep-learning classification, 3D grading and continuous learning run as software that updates and integrates long after install.
WebClassifier classifies and grades every defect by type and severity, at over 99% accuracy.
Automatic, optimal inspection settings stored per fabric style.
Continuous full-length edge / centre / edge shade-variation measurement over the whole web. Enables roll shade sorting.
Self-validation proves the system is catching what it should.
Maximise yield against customer and product rules, recalculated on demand.
Drives re-roll machines, rolling, stopping and cut positioning, precisely.
Regulate and control quality consistently across multiple supply chains.
Classification accuracy builds over time as more fabric is processed.
Customer results
From heritage check designs to safety-critical automotive textiles and high-volume finishing.
Heritage check patterns defeat conventional vision systems. WebSpector, combined with deep-learning AI WebClassifier, separates the design from the defect, catching genuine faults across complex repeats with minimal false alarms.
Safety-critical automotive textiles allow no escapes. End-to-end inspection, raw web to final cut, delivered consistent, auditable quality across the supply chain.
Inline systems integrated with an optimised cut plan deliver 2–3% more first-quality yield versus manual inspection, with less rework, waste and embodied carbon.
The investment in the Shelton state-of-the-art vision system marked a major step forward in our company’s commitment to operational excellence. By replacing traditional manual inspection methods with automated, data-driven solutions, the system has enhanced both the accuracy and consistency of quality control across our production processes.
Our inspection is now fully automated with just a quick commercial review of the results, then creation of an optimised cut plan which is performed on automated roll, cut and pack machines.
We conducted trials on most of the systems on the market and concluded that the Shelton Vision system was the only one that could inspect our products without drowning our operators in false alarms.
The investment in the system aligns with our strategic focus on value creation, sustainability, and delivering high-performance textile solutions to global markets. This technology represents a step-change in how we manage quality. By embedding automatic inspection into our process, we are not only improving yield and reducing waste, but also strengthening the consistency and reliability our customers expect.
Sustainability
Every metre of fabric rejected carries the carbon of every process that made it, fibre, dyeing, finishing. By catching instability early, WebSpector prevents embodied-carbon waste instead of compensating for it after production. Connected to plant energy and resource data, inspection becomes verified environmental monitoring at fabric-roll level.
Applications
WebSpector runs wherever fabric quality is critical and waste is costly.