Changeover from 45 to 12 Minutes: How an Engineering Company Cut Costs Without Layoffs

Client

An engineering company with 240 employees, a supplier of machined parts for the automotive and energy industries. Three shifts, a machine park of CNC lathes and milling centres, and a growing share of small batches — typically 50 to 400 pieces per order. The company is presented anonymously; this is a composite case study drawn from our projects in this segment.

The project was initiated by the CFO after the margin of the small-batch segment had dropped by four percentage points year on year — and nobody in the company could say precisely why.

Challenge

Customers were pushing for smaller batches and shorter lead times, and the number of changeovers per week had almost doubled in two years. A single changeover took 45 minutes on average — and because it was done "on the fly", missing fixtures and gauges were being hunted down while the machine stood still. The consequences stacked up: the schedule was rescued through weekend overtime, the scrap rate after changeover reached 4.2%, and the margin on small batches was approaching zero. Management was considering turning small orders away — precisely the segment that was growing.

At roughly 60 changeovers per week, 45 minutes per changeover meant over 40 hours of pure machine downtime — the equivalent of a full shift of the entire hall. The company also had one internal improvement attempt behind it: a changeover schedule on the notice board and an appeal for discipline. It lasted six weeks. The setters were not the problem — each had his own routine, refined over years; what the company lacked was one shared routine, measured and proven.

Approach

1. Data analysis and video recording of changeovers (2 weeks). We filmed 18 changeovers on three representative machines and reviewed them with the team minute by minute. The finding: 60% of the time consisted of activities that could have been done while the previous order was still running — searching for tools, preparing fixtures, waiting for the crane. The videos also served as a common language — instead of guessing who was slow, the team talked about specific minutes.
2. SMED workshop with supervisors and setters. Together we split the activities into internal (machine must be stopped) and external (can be done in advance). Each machine received a one-page changeover standard with the sequence of steps and a first-piece inspection. The first-wave target was deliberately conservative — 20 minutes; more ambitious goals came only after the first successes.
3. Workplace preparation. Changeover trolleys were introduced, carrying the complete set of tools and fixtures for the next order and prepared while the machine was still running. Gauges and measuring equipment were moved to the machines, each with its own labelled position.
4. Training and daily management. The setters were trained on the new standard, changeover times were recorded daily on the team board and reviewed once a week in a 15-minute meeting. Deviations were addressed immediately, not at the end of the month. After the third month, the production director took over these meetings and the project ran on without our daily presence.

Results

Five months after the project started:

• Changeover time: from 45 to 12 minutes on average across the monitored machines (−73%), with no investment in new machinery — the most expensive item was the changeover trolleys.
• Scrap rate after changeover: from 4.2% to 2.6% (−38%), thanks to the first-piece inspection standard.
• Weekend overtime in production fell by two thirds; the freed-up capacity absorbed the growth in small batches without turning orders away. Payback period of the entire project: 5 months.

Worth noting is also what did not happen: no layoffs, no purchases of new machines, no small orders turned away. And the system lives on — the setters who co-created the standard updated it twice on their own over the following year.

The same principle — first measure where the money is actually leaking, then change the processes, and only then consider investment — is the basis of all our cost reduction projects. If your margin is disappearing into small batches, downtime or scrap, we can show you fairly quickly where to start.

29.07.2026