This article is for quality managers, production managers, continuous improvement leaders and plant managers who open nonconformance reports every week and keep seeing the same defects come back. It addresses one specific question: how to link each nonconformance report to what the machine was doing just before the defect (a stop, a restart, a changeover, micro-stops, a slowdown), so that root cause analysis starts from measured facts rather than from memory.
- A nonconformance report describes the part, rarely the state of the machine that made it.
- Restarts, changeovers and micro-stops often come just before scrap.
- A precise timestamp lets you pull the machine window before the defect.
- Reading scrap and stops together leads to better corrective actions and OEE.
A well-structured report that often lacks machine context
Lumivi, a Lyon-based firm that builds custom business tools and provides consulting for small and mid-sized companies, published Fiche de non-conformité : modèle, champs et exemple ISO 9001 (in French) on its blog, written by co-founder Loïc Mabilon. The article details the fields of a nonconformance report (identification, finding, disposition, cause, corrective action), points out that ISO 9001 does not impose a standard form but does require this information to be retained, and provides an Excel template and a completed industrial example. It stresses a point many shops neglect: “Le constat est rempli par la personne qui détecte l’écart, au moment de la détection.” (in other words: the finding is filled in by the person who detects the deviation, at the moment of detection).
That principle of immediate recording is the right foundation. One limit remains that even the best form cannot remove on its own: the person who detects the defect sees the part, not necessarily what happened on the machine in the minutes or hours before. Yet a large share of scrap starts with a machine event: a restart after a stop, the first parts after a tool change, a run of micro-stops, a drifting cycle rate. Lumivi’s article structures the report; this one looks at how to enrich it with measured machine history.
What the machine knows at the moment of the defect
When a part is declared nonconforming, the useful question is not only “what is the defect?” but “what state was the line in when this part was made?”. The following events are the ones most often linked to a defect, and each one completes a field of the report.
| Machine event before the defect | What it may explain | Report field to enrich |
|---|---|---|
| Restart after a long stop | Parts made while warming up, settings still stabilizing | Finding: position of the part after the restart |
| Changeover or tool change | First parts out of tolerance, incomplete setup | Identification: product, tool, time of the changeover |
| Repeated micro-stops | Unstable feeding, jams, poorly adjusted sensor | Cause: micro-stop frequency over the period |
| Cycle rate slower or faster than nominal | Process pushed or held back outside its window | Cause: rate deviation at the time of the defect |
| Unexplained stop | Untracked intervention, undocumented adjustment | Disposition: question to ask the team concerned |
None of these events proves the cause of the defect on its own. They give a verifiable, timestamped lead that prevents jumping to conclusions. Lumivi’s article is clear on this: operator error is not a root cause. Machine history helps you look further.
Linking a nonconformance report to machine history, step by step
- Timestamp the finding. The date alone is not enough: record the time of detection and, if possible, the production time or batch of the part. That is the key to finding the state of the machine.
- Identify the machine and the run. Machine, workstation, part number, tool fitted. Without this, the report stays attached to a product, not to a process.
- Pull the machine window before the defect. Look at the period before the part was produced: stops with duration and declared reason, micro-stops, changeovers, actual cycle rate. The width of the window depends on the process: for example, on a press a team might look at the preceding hour; on a furnace, the length of the thermal cycle.
- Record the context in the report. Add a “machine context” field: one factual line, for example “part produced shortly after a restart following a 40-minute stop” (illustrative example). No interpretation, only facts.
- Look for repetition. A single report says little. Ten reports that all fall on parts made after a restart point to a serious lead for corrective action.
- Check the action worked. After the corrective action, compare scrap in the same situations (after restart, after changeover) before and after. The report closes on a measurement, not on an impression.
Scrap, the OEE quality factor and hidden losses
OEE breaks down into availability, performance and quality. Nonconformance reports deal with quality, stops with availability, slowdowns with performance. In many plants these three topics are tracked by different teams in different files. The link between a stop and the scrap that follows it gets lost.
Connecting them changes the picture. For example, a 20-minute stop does not cost only 20 minutes: it also costs the parts scrapped at restart and the time needed to get back to steady running. A poorly set changeover weighs on both availability and quality. Micro-stops, invisible on manual logs, often show up just before a quality drift. When you look at scrap alongside machine history, you choose projects better: sometimes the most profitable quality lever is a restart procedure, not one more inspection at the end of the line.
What TEEPTRAK adds to the nonconformance report
TEEPTRAK measures OEE, stops and micro-stops in real time. A module wired into the electrical cabinet and a sensor at the line record machine activity automatically, with no connection to the PLC. Installation takes about one hour per machine and OEE is available within 48 hours. Every stop, micro-stop and rate change is timestamped, so the quality manager can find the state of the line at the moment a nonconforming part was made within minutes. The report remains the tool of the quality process; TEEPTRAK provides the machine history that feeds it. The solution is deployed in more than 700 plants in over 30 countries.
Further reading
- Fiche de non-conformité : modèle, champs et exemple ISO 9001 (Lumivi, in French)
- Quality Rate OEE: How to Calculate, Measure and Improve This Critical Manufacturing Metric
- Before SMED: measure every changeover to decide where to start
FAQ
Should a nonconformance report be opened for every restart scrap part? Not necessarily. If restart parts are expected and covered by a procedure, they are logged as startup scrap. If they exceed what is expected or the defect is unusual, a report is justified, with the machine context.
How far back should you look before the defect? There is no universal duration. Start from the process: the time it takes for a machine event to show up in a finished part. Set that window in writing, per machine, then adjust it with use.
Does machine history replace root cause analysis? No. It provides timestamped facts that guide the analysis and prevent jumping to conclusions. The root cause is always confirmed on the shop floor, with the team that knows the process.
0 Comments