Plant floor data joined to the ERP at last
Line telemetry and ERP had never been connected. We modelled both into one layer so OEE, scrap and maintenance all read from the same numbers.
- OEE
- +9%
- Lines connected
- 14
- Scrap reduction
- -12%
- Time to first value
- 8 wks
The situation
Fourteen lines, a historian recording every tag, and an ERP holding every order. Neither system knew about the other. OEE on the shop floor screen and OEE in the monthly pack disagreed by several points, and both were defensible.
Constraints
The plant could not take downtime for the work, and the maintenance team had no appetite for another dashboard they had not asked for.
What we did
We modelled historian data and ERP into one layer, so throughput, scrap and downtime reason codes all resolve to the same shift and the same order.
Rather than build a new interface, we fed the existing shift handover screen. The team kept their routine and the numbers underneath it changed.
What came next
With a trustworthy year of history, predictive maintenance on two bottleneck machines became worth doing. That was scoped as a separate engagement, after the foundation proved itself.
Results
| Measure | Before | After |
|---|---|---|
| Overall equipment effectiveness | 68% | 77% |
| Scrap rate | 4.1% | 3.6% |
| Lines reporting live | 0 | 14 |
| Shift handover | Clipboard | Dashboard |
What we left behind
- A project in the client's own repository, with tests and CI
- Documented definitions with named owners
- Runbooks for every scheduled job, written for a non-engineer
- Two weeks of training for the team taking it over
Other engagements you might want to read
Tell us which number you cannot trust
Send one paragraph about the report that always starts an argument. A data lead, not a salesperson, replies within one business day.