Method · SMED changeover reduction

SMED and Changeover Reduction: The Scoped Sprint That Frees Capacity

How a SMED sprint works: film a real changeover, split internal from external work, standardize, and hold the gain. Formats, costs and the GMP variant.

Short answer

A SMED sprint reduces changeover time in a fixed sequence: film a real changeover, classify every task as internal (machine stopped) or external (can happen before or after), move and parallelize the external work, standardize the new sequence, and only then consider technical modifications. Typical first-sprint results are 30 to 50% less changeover time, mostly from organization rather than engineering.

Updated: 2026-08-07 · Reviewed: Lean Competence practice editors

High-mix plants pay for changeovers twice: once in the lost hours themselves, and again in the big batches scheduled to avoid them, with all the inventory and inflexibility big batches drag along. SMED, single minute exchange of die, is the fifty-year-old method that attacks both payments at once, and it remains one of the highest-return scoped engagements in manufacturing because most of the gain requires organization, not investment.

How the sprint actually runs

Film a real changeover. Not the demonstration version: a normal one, with the normal crew, on a normal day. The video is the dataset, and it is regularly a revelation; teams watching their own changeover discover walking routes, searching, waiting for the crane, waiting for quality, waiting for the missing gasket.

Classify every task. Each activity is tagged internal (only possible while the machine stands) or external (can happen before the stop or after the restart). First-time analyses typically find 30 to 50% of stopped-machine time spent on work that never needed the machine stopped: fetching tools, preparing materials, filling in paperwork.

Move, parallelize, eliminate. External work moves out of the stop window and into preparation checklists. Internal work gets parallelized across crew members with clear roles, and its worst steps get simplified: quick-release instead of bolts, preset instead of adjust, standard heights instead of shimming. Technical modifications come last, after the organizational gains, because half of the equipment wish list usually evaporates once the sequence is clean.

Standardize and hold. The new changeover becomes standard work: sequence, roles, preparation checklist, target time, and a visible record of every changeover against target. This step decides whether the sprint bought a permanent capacity increase or an anecdote. Variation between shifts is the tell: if Tuesday night takes twice as long as Monday morning, the standard is decoration.

A focused sprint runs three to five days on site plus preparation and a follow-up visit, priced in Switzerland at roughly CHF 25'000 to 60'000 as a scoped engagement depending on line complexity and regulation (market context here). First sprints on lines that never had structured changeover work typically deliver 30 to 50% reduction, and as important, a halved spread between best and worst changeovers.

The regulated variant

On GMP lines, changeover includes cleaning, line clearance, documentation and release, and those steps often dominate the mechanical work. The method holds, but the classification gains a dimension: what is validated and untouchable as-is, what is procedural and adjustable through document control, and what is organizational and free. Big wins hide in parallelizing documentation with physical work and in preparation checklists that prevent the clearance-failed-restart loop. We wrote up the field experience separately in SMED on regulated lines, and the sector context in the lean pharma guide.

What changeover reduction buys

The obvious return is capacity on constrained equipment: hours per week recovered without capex. The strategic return is smaller batches becoming economical, which shortens lead times, cuts inventory and makes the plant honest about demand instead of producing to changeover-avoidance. Plants that halve changeover times and then leave batch sizes untouched have collected half the prize.

FAQ

What does SMED stand for?

Single Minute Exchange of Die, the historical target of under ten minutes. The name is legacy; the method is the internal-external discipline, and it pays off long before any single-digit result.

How much changeover reduction is realistic?

First structured sprints typically achieve 30 to 50%, dominated by moving external work out of the stop window and parallelizing crew roles. Later iterations with technical measures go further.

How long does a SMED engagement take?

Three to five days on site for the sprint, plus preparation and a follow-up visit some weeks later to audit the standard against reality. Scoped cost in Switzerland: roughly CHF 25'000 to 60'000.

Does SMED work on packaging and filling lines under GMP?

Yes, with the validation-aware variant: cleaning and release steps are classified alongside mechanical ones, documentation is parallelized, and anything validated routes through change control. Reductions of a third are common without touching validated parameters.

Field notes, in your inbox

One email,
every other Friday.

What our practitioners are seeing on real shop floors, plus the latest episode and the next event. No marketing, no recycled lean LinkedIn posts.

EMAIL

4'200+ subscribers

· Unsubscribe anytime