Interview with Marc Ewald, Managing Director of Liersch, Part 2

In your experience, which sewing defects are still underestimated by manufacturers?

Two in particular. The first is skipped stitches. A thread breakage is obvious – the machine stops, somebody fixes it. A skipped stitch is silent. The part looks finished, it passes visual inspection, and the weakness only shows up under load, sometimes at the customer, sometimes years later.
The second is drifting thread tension. It rarely fails suddenly; it moves slowly, and because it moves slowly the operator adapts to it without noticing. The seam still looks acceptable but the mechanical properties are no longer what the specification says.
Both belong to the same category: defects that are invisible in the finished product. That is why we built TMS around continuous monitoring of the process rather than inspection of the result – you cannot inspect quality into a seam after the fact.

Discover how the Liersch Thread Monitoring System helps detect critical deviations in the sewing process and supports reliable seam quality.

https://liersch.com/online-shop/eigene-produkte/fadenueberwachung-tms/ 

Many seam defects are only discovered during final inspection—or even after delivery. Why is this such a challenge?

Because the cost of a defect multiplies with every step it survives. Caught at the machine, it costs a few seconds. Caught at final inspection, it costs the value already added to the part plus rework. Caught at the customer, it costs a complaint, a containment action and credibility. And in safety-critical applications – restraint systems, PPE, lifting equipment – the worst case is not a commercial one at all.
There is a second problem: traceability. If you find a defect at final inspection, you rarely know when the process started drifting, so you have to quarantine everything back to the last known good point. Real-time monitoring turns that open question into a defined answer – you know which part, which seam, which moment.

Your Thread Monitoring System (TMS) enables real-time monitoring of the sewing process. What problems does it solve that traditional monitoring cannot?

Conventional monitoring is essentially a thread-break detector: it reacts once the thread is gone. TMS looks at the process while it is still running. It measures the behaviour of the thread continuously, stitch by stitch, and detects deviations before they become visible defects – skipped stitches, tension drift, changes in thread consumption.

Three things follow from that. First, the machine stops at the right moment, so scrap stays at one part instead of a whole batch. Second, every seam produces a data record, which means real traceability instead of sampling. Third, the data can be used beyond quality assurance – thread consumption per part, machine availability, patterns that point to maintenance needs.

It is also worth saying that TMS is not a lab project. It was developed within a publicly funded cooperation project and was recognised as ZIM Cooperation Project of the Year 2026 by the Federal Ministry for Economic Affairs and Energy – and it is in series use in demanding production environments.

A monitoring system is only as reliable as the process it monitors. How important is the sewing thread itself for obtaining consistent and meaningful process data?

It is decisive – and this is the point where our work with AMANN becomes very concrete. A monitoring system measures deviations from an expected behaviour. If the thread itself varies from lot to lot, in friction, in lubrication, in evenness, then the system sees variation that has nothing to do with the process. You either lose sensitivity or you generate false alarms, and a system that cries wolf gets switched off by the operator within a week.

A premium thread with consistent, predictable properties gives us a stable baseline. That is what makes tight monitoring thresholds possible in the first place. Put simply: thread quality does not only determine seam quality – it determines how much you can learn from the process at all.