Customer story
Industry
Aluminium extrusion and anodising
Size
240 employees, € 68 mln revenue (2025)
Location
Kortrijk, Belgium (two press lines)
Using since
8 Sep 2025
−38%
Average changeover time
74 → 46 minutes, timed die-out to first good profile across 1.248 changeovers.
−53%
Overtime on the press floor
312 → 148 hours per month; no Saturday shift called since April 2026.
+6,8pp
Equipment effectiveness, P3
OEE 61,2% → 68,0% on the busiest of the two presses, same product mix.
We run about 1.400 profile dies across two presses. From 2024 our order book tilted away from long runs for a handful of large fabricators towards short runs for many smaller ones: average order size fell by roughly a third, so the number of die changes rose faster than anything else in the plant.
The week was planned every Friday in a spreadsheet, then re-planned twice a day by whoever was on shift. Jobs were sequenced by promised date, never by tooling similarity, so we would swap from a coated colour to a mill finish and back again and pay the full cleaning cycle each time. About a fifth of our available press hours went to changing dies rather than extruding.
Everyone on the floor knew the answer was to group similar changeovers. Nobody had the time to work out what similar meant across 1.400 dies, twice a day, with three-quarters of the order book moving every week.
We had run SMED training twice before and both times it faded within a quarter. What changed is that the schedule stopped fighting us — the plan now reflects how the presses actually behave on a Tuesday afternoon.
Ilse Vandenberghe · Operations Director, Verhaeghe Profielen NV
The approach, the full before-and-after table and the scope of work follow on page 2.
Nothing was installed in the first month. We timed changeovers with the die-setters, agreed one definition of what a changeover is, and only then let a scheduler touch the sequence.
Two of the three steps below are now run by the plant itself; we only join the monthly review.
Two weeks on the floor, no software
We stopwatched 46 changeovers on both presses with the setters and split every minute into internal and external work. That baseline — 74 minutes average, 61,2% OEE on P3 — is the “before” column of this case study.
A similarity score for 1.412 dies
Die geometry, alloy, cavity count and surface treatment were tagged from the ERP export, so any two consecutive jobs could be scored on how much of the change they share. Five weeks of work, three of them cleaning the tooling master.
Sequencing handed back to the shift
Shift leads now approve a proposed sequence twice a day instead of building one from scratch. An override takes about four minutes and the model re-scores everything downstream, so the plan survives a rush order instead of being abandoned.
The table compares H1 2026 with H1 2025 on the same two presses, the same product-mix definition and the same measurement — die-out to first good profile. Both presses run three shifts, five days: 240 press hours a week. Changeovers now absorb 15% of those hours instead of 21%, while the plant runs 17% more of them.
Two caveats we would rather state: P2 was re-motored in November 2025, which we estimate accounts for 4 of the 28 minutes saved, and the overtime figure benefits from a milder order peak in Q2 2026.
| Metric | Before · H1 2025 | After · H1 2026 | Change |
|---|---|---|---|
| Average changeover time | 74 min | 46 min | −38% |
| Overtime, press floor | 312 h / month | 148 h / month | −53% |
| OEE, press P3 | 61,2% | 68,0% | +6,8pp |
| Changeovers per week | 41 | 48 | +17% |
Sequent Flow scheduling module (2 presses, 12 named users) · two-week on-site SMED baseline · tooling-master enrichment for 1.412 dies · six shift-lead sessions · monthly review for the first two quarters.
Live on P2 from 8 Sep 2025, on P3 from 20 Oct 2025. No change to the ERP.
The same sequencing logic moves to the anodising line in Q4 2026, where colour changes carry a comparable cleaning penalty.
A tooling-condition signal is in pilot, so dies approaching maintenance are scheduled next to a planned stop instead of a rush order.