How standardised OEE and downtime measurement across every line at a regulated manufacturing site — so the plant stopped arguing about the data and started fixing the losses behind it.
At a glance
- Client: Regulated manufacturing site (multi-line)
- Sector: Regulated production — quality and traceability critical
- Problem: Each line measured downtime and OEE differently, so nobody trusted the numbers
- Scope: OEE Foundation training, KPI and downtime definition, reason-code design, data-quality review
- Result: One consistent, trusted OEE standard across the whole site
- No new hardware: Delivered by fixing the measurement system and the behaviour on the floor
The problem: the meeting started with an argument
The site was already collecting production data. That was not the issue. The issue was that no two lines counted it the same way.
One line started the downtime clock the moment a machine stopped. Another waited until the operator had tried the obvious fixes. One team logged a changeover as planned time. The next team logged the same event as unplanned. When two lines reported the same OEE figure, it did not mean the same thing.
So every performance meeting opened the same way. Operators questioned the figures. Managers were looking at numbers that did not line up. Before anyone could talk about a loss, they had to settle a debate about whether the loss was even real.
If people do not trust the number, they will not use it. And a number nobody uses cannot drive an improvement.
What was done?
We worked directly with the production teams to build one way of measuring and using OEE, and to make it stick. The work covered the parts of the measurement system that were quietly disagreeing with each other:
- Clear OEE definitions, written in plain language
- Consistent downtime rules — when the clock starts, when it stops
- A practical reason-code structure operators can actually use mid-shift
- Common production standards so the same job runs to the same rate everywhere
- Clear rules for who captures data, and when it gets reviewed
- Operator and supervisor training through OEE Foundation programme
The training did more than explain what to enter. It explained why consistent data matters — so operators understood that the reason code they pick at 2am is what the improvement team reads the next morning. When the people entering the data see the point of it, the data gets better on its own.
What changed on the floor
Once the rules were agreed and trained, the same production event was recorded the same way on every line. The debates that used to eat the first ten minutes of every meeting simply stopped happening.
| What used to happen | What happens now |
| Downtime start and stop was judged differently on each line | One written rule for when the downtime clock starts and stops |
| Operators picked reason codes inconsistently | A fixed, practical reason-code structure everyone uses |
| Planned vs. unplanned time was interpreted case by case | Planned and unplanned time are defined and agreed |
| OEE was calculated in more than one way | A single OEE calculation across the site |
| Line-to-line comparison was not trusted | Lines can be compared on a like-for-like basis |
| Meetings began by debating the data | Meetings begin with the biggest loss |
The measurable side of trust
Better definitions change how a plant spends its attention. Before standardisation, roughly 40% of the daily meeting went on arguing about the data; afterwards, closer to 5%, leaving about 95% of the meeting for what to actually improve. (Illustrative — confirm with your own before-and-after once the standard has run for a quarter.)
The result: one trusted view of the plant
- Daily performance meetings that start on the real problem
- Line-to-line comparison that holds up
- Downtime analysis that points at genuine losses
- Improvement priorities everyone accepts
- Management reporting that matches what the floor sees
The Approach
Reliable OEE is not a dashboard problem. It is a definitions-and-behaviour problem. We fixes it in five steps, in this order: 1) Define the rules, 2) Train the people, 3) Measure consistently, 4) Build trust in the data, 5) Use it to improve.
How we can support
- OEE Foundation training
- KPI and OEE definition
- Downtime structure
- Reason-code design
- Operator and supervisor training
- Data-quality reviews
- Daily management routines
- Cross-line standardisation