This article is for plant managers, production managers, continuous improvement and operational excellence leads, industrial engineering and maintenance managers who have already rolled out routines, standards and problem solving tools, and who see that the results do not hold. It explains why an operational excellence approach designed as a system needs reliable, shared measurement of downtime and OEE, taken at the machine, so that these tools work on facts rather than impressions.
- Operational excellence tools turn information into decisions.
- If downtime data is incomplete or disputed, routines and standards run on empty.
- Reliable measurement is automatic, qualified on the floor and the same for everyone.
- Measure on a narrow scope first, then plug the data into the routines.
What GXG Consulting says about tools that are no longer enough
GXG Consulting, the French operational excellence consultancy led by Gilles Galichet, published the article Excellence opérationnelle systémique : pourquoi les outils ne suffisent plus (et ce qui fonctionne vraiment) (in French). Its starting point is blunt: plants have practised Lean, Kaizen, 5S, TPM or Six Sigma for years, yet performance stays unstable and gains fade once external support ends. For GXG Consulting, a tool sits inside an existing system (implicit rules, decision making, management routines) and, if that system does not change, it absorbs the tool: standards get bypassed and indicators become decorative. The firm proposes a five pillar model, the last of which covers performance steering and continuous improvement, with a line that sums up the issue: “L’essentiel est de rendre les écarts visibles et traitables.” (“What matters most is making deviations visible and actionable.”)
That sentence is the starting point of our article. Making a deviation visible means seeing it when it happens, naming it the same way from one shift to the next and expressing it as lost time. This is where measurement comes in: not as one more tool, but as the common data the other tools rely on.
Without reliable measurement, every tool runs on empty
A routine, a problem solving session or a standard produces nothing on its own. It turns information into a decision. If the information is incomplete, late or disputed, so is the decision. The table below shows what each tool expects from measurement, and what happens when it is missing.
| Tool | What it expects from measurement | What happens without reliable measurement |
|---|---|---|
| Daily production meeting | Yesterday’s stops, with their duration and cause, available before the meeting | The meeting discusses impressions and spends its time rebuilding the facts |
| Problem solving (QRQC, 8D, PDCA) | A machine’s downtime history, its frequency and its split by shift | The team tackles the problem it remembers best, not the one that costs the most |
| Work standards | Evidence that a deviation from the standard turns into stops or slowdowns | The standard becomes a rule with no visible consequence, then it gets bypassed |
| Indicators and dashboards | An OEE calculated the same way on every machine and for every shift | The indicator is argued about instead of used, and ends up decorative |
| Maintenance (TPM, autonomous maintenance) | Micro-stops and repeated drifts, even very short ones | Only major breakdowns are seen, small losses stay in the noise |
The last column matches what GXG Consulting describes when a tool is absorbed by the system. Measurement alone does not fix these drifts, which also depend on governance and management. It does remove a common excuse: not knowing what really happened on the machine.
What “reliable and shared” means at the machine
Reliable measurement is not measurement that looks more precise on paper. It is measurement that depends neither on memory nor on the goodwill of whoever enters it. Three conditions matter in practice.
- It is automatic. The machine state (running, stopped) is captured directly, with no manual entry. Short stops, which nobody writes down, are counted too.
- It is qualified by the people who know. The operator or shift leader gives the cause of the stop, from a short list shared by all shifts. The machine says when and for how long, people say why.
- It is the same for everyone. Operators, shift leaders, maintenance, engineering and management look at the same figures, calculated with the same rules. Nobody walks into the meeting with their own spreadsheet.
“Shared” is probably the key word. For example, on a line where the morning shift logs a stop as “adjustment” and the afternoon shift logs the same stop as “breakdown”, no routine can settle the matter, because the discussion is about the data, not the problem. A common list of causes and a duration captured by the machine bring the discussion back to what matters: what do we do so that it does not happen again?
A five step method to put measurement to work for the tools
The approach below starts at the machine and works up to the routines. It does not replace work on vision, governance or skills; it gives that work a factual base.
- Pick a narrow scope. One line or cell, ideally the one where downtime is argued about most often. The goal is to learn fast, not to cover the whole plant.
- Capture stops automatically for a few weeks. Without changing existing routines, compare what the machine records with what was reported. The gap between the two is often the first useful result.
- Build a short, common list of causes. With operators, shift leaders and maintenance, agree on a list everyone understands the same way and can fill in within seconds.
- Plug measurement into the routines. The daily meeting starts from yesterday’s stops, problem solving starts from the Pareto of causes, and the standard is revised when the data shows a repeated deviation.
- Check that the loop closes. For each action decided, look over the following weeks at whether the targeted stop actually goes down. An action with no measurable effect is reviewed, not forgotten.
TEEPTRAK provides this measurement. A module wired into the electrical cabinet and a sensor at the line record every stop and micro-stop automatically, with no PLC connection. Installation takes about one hour per machine and OEE is available within 48 hours. Operators qualify causes on the spot and every shift sees the same data. TEEPTRAK is used in more than 700 plants in over 30 countries. Measurement replaces neither method nor management: it gives them facts.
Further reading
- Excellence opérationnelle systémique : pourquoi les outils ne suffisent plus (et ce qui fonctionne vraiment) (GXG Consulting, in French)
- Daily production meeting: 15 minutes with yesterday’s downtime data
- QRQC with machine data: downtime and scrap facts for faster problem solving
FAQ
Why don’t operational excellence tools always deliver lasting results? Often because they are laid on top of a system that does not change and rely on disputed data. Without reliable, shared downtime measurement, routines discuss impressions and standards lose their link with actual performance.
What is reliable OEE measurement? Measurement captured automatically from the machine state, which also counts short stops, with causes qualified from a list shared by all shifts. It is calculated with the same rules on every machine.
Should you measure before or after setting up the routines? Both move forward together. Measuring for a few weeks before changing the routines gives a starting point and reveals the gap between reported and actual stops; the routines then use this data as soon as it is available.
0 Comments