Augos Factory / Learn / SMED and changeovers
Learn · lean changeovers
Your fastest changeover is your target.
In short
SMED (single-minute exchange of die) is a lean method for cutting changeover time. Work is moved from internal (line stopped) to external (line still running), then what remains is streamlined and standardised. The most credible first target is the fastest changeover your own team has already done.
The method
The five steps of SMED
Developed by Shigeo Shingo for press tooling, SMED now applies to any changeover. Think of a flavour change on a filler, a film change on a wrapper or a mould change on an injection press.
Measure the changeover as it is
From the last good unit of one product to the first good unit of the next, at full speed.
Separate internal from external work
What truly needs the line stopped, and what could be done while it runs?
Move internal work outside
Pre-stage tools, parts and materials; pre-heat; pre-set.
Streamline what remains
Quick-release fixings, fewer adjustments, parallel work.
Standardise and check
Write the method down, train it, and measure the changeovers in the weeks after, at least ten.
One changeover through the five steps
IllustrativeIt is the pit-stop principle. The tyres, tools and crew are ready before the car comes in, so the car stands still only for the work that needs it stopped.
Illustrative changeover, not a customer.
Variation
Same changeover, same line: why seven minutes one day and forty the next?
If the same changeover varies that much, the fast version is already possible. Your team has done it. The question is what was different on the fast days: which team, which product sequence, which tools were ready.
That is why the target comes from the line's own recommended time rather than a textbook figure. Variation and standard work ›
Changeovers against the allocated time, by team
IllustrativeThe same changeover on the same line: six teams, 30 days, 30 changeovers.
Seven minutes at best, forty at worst. 16 of 30 overran the allocated time, yet a quarter of them took 12 minutes or less. Team C's way of working is the one to write down.
Illustrative changeovers, not a customer.
Measuring changeovers
Allocated time plus overrun
Every changeover is split two ways. The allowed time is planning's (Planning Efficiency Loss); the overrun beyond it is the operator's (Operator Compliance). It stays a planned stop however long it runs. The line's own best is the benchmark: where it beats the allocated time, the allocated time can be tightened.
One changeover, split two ways
IllustrativeIllustrative stop, not a customer.
Run SMED on the record
Every changeover measured against your own best
Step one of SMED measures the changeover as it is. Step five checks that the new method held. The Planned Stoppage Analysis does both, on every changeover, for every crew.
Each changeover is measured against its allocated time, the typical time, the recommended time and the team's best. Switch between them and the recoverable minutes change, so the team knows which target it is chasing.
Recommended is the median less the median absolute deviation of the line's changeovers over the calendar year to date, once there are ten. It is a fast pace the line already reaches regularly. When crews beat the allocated time, planning can tighten the allocated time, and the next gap is measured from there.
Planned Stoppage Analysis · one changeover · 30 days
Illustrative| Crew | Changeovers | Median | Within allocated | Recoverable against recommended |
|---|---|---|---|---|
| Crew A | 5 | 13.5 min | 4 of 5 | 9.5 min |
| Crew B | 5 | 17 min | 2 of 5 | 32 min |
| Crew C | 5 | 10 min | 4 of 5 | 4 min |
| Crew D | 5 | 19 min | 1 of 5 | 42.5 min |
| Crew E | 5 | 21 min | 1 of 5 | 32 min1 outlier left out |
| Crew F | 5 | 15.5 min | 2 of 5 | 17.5 min |
| All crews | 30 | 15.8 min | 14 of 30 | 137.5 min |
Crew C's median is 10 min, under the Recommended 12. Its method is the one to write down in step five, then train to every crew.
Illustrative data, not a customer. The same 30 changeovers as the chart above, laid out as the Augos Planned Stoppage Analysis. Stops above the maximum threshold are treated as outliers. Here the benchmarks come from these 30 changeovers; on the platform they use the calendar year to date.
In the toolbox
Questions
Changeovers, answered
When does a changeover start and end?
From the last good unit of the old product to the first good unit of the new one at full speed. Run-up time counts.
How do you set a changeover's allocated time?
From the line's own recommended time for that changeover reason, over the calendar year to date. Tighten it as the team improves.
Is SMED only for big presses?
No. The same steps cut flavour changes, film changes, mould changes and cleaning between products.
Related
Lean manufacturing in the knowledge base › Where SMED sits among the Lean methods, and what the platform measures for each.
See SMED on your line.
Thirty minutes with a manufacturing specialist: each changeover against its allocated time, and the fastest your team has already done.
- Every changeover timed
- Allocated time and overrun split
- Prices in rand, ex VAT
A manufacturing specialist replies within two business hours to set a 30-minute call.