Experience Is Retiring: Knowledge Transfer in Manufacturing When the Boomers Leave

Every plant has that one person. The setter who hears when a tool needs retightening. The maintenance technician who knows this line starts differently in winter. Nobody ever trained them in it, and they never wrote it down. They simply know.

And one day they work their last shift.

What happens next is often described most precisely by people in safety-critical work — air traffic control, for instance, where the gap between standard procedure and experience is unusually visible.

It is not positions that leave, it is experience

The figures behind this became sharper in July 2026. The Skilled Labour Competence Centre at the German Economic Institute counted them: 8.5 million employees in Germany subject to social insurance contributions are 55 or older. That is 25.4 percent — a good quarter of the entire workforce, leaving for age reasons within the next 10 to 15 years.

What matters is not the headline number but its composition. 4.6 million of those employees work at skilled-worker level. 3.6 million work in occupations that are already short of people.

An earlier analysis put the gap in plain terms: around 4.5 million skilled workers are leaving, against roughly 3.2 million skilled workers under 30 coming in. That difference is no longer workforce planning. It is a structural break.

The loss is therefore qualitative as well as numerical. Someone who leaves after decades takes with them not just a position but the ability to resolve difficult situations quickly — situations that take inexperienced colleagues considerably longer.

Mentoring works — but right now nobody has time for it

Here is the real contradiction, and it is rarely stated plainly.

The most effective known method of passing on experience requires two people spending time together: someone experienced who demonstrates, and someone inexperienced who follows along. That time is exactly what the labour shortage destroys.

Germany's Institute for Employment Research measured how long it takes to fill a vacancy: 70 days in 2010, 95 days in 2023. The vacancy overhang rate for qualified workers in western Germany recently stood at 45.6 percent — for almost every second opening, no suitably qualified person is regionally available.

The German Chambers of Commerce report that over 50 percent of surveyed companies struggle to fill positions, particularly in the industrial sector. In the metal and electrical industry, more than 110,000 skilled workers were missing in a single twelve-month period.

The result is a loop. The workforce shrinks, those remaining absorb more, and precisely because of that there is no time to train the newcomers meant to close the gap. The people who ought to be demonstrating have to produce instead.

Translating work instructions does not solve the language problem

A second finding has long been everyday reality on the shop floor. At the end of 2023, over 27 percent of Germany's population had a migration background, and companies name language gaps as the single most significant barrier to hiring — even for well-qualified specialists.

The obvious answer is to translate the work instructions. Research considers that insufficient:

The reason is the same one that makes tacit knowledge hard to capture: a translated sentence carries the words, not the movement. Someone who has never seen how tight "hand-tight" is cannot read it up in any language.

What a publicly funded research project found

That this is not a niche problem is evident from the effort spent on it. Between August 2022 and January 2025, a consortium funded by Germany's Federal Ministry of Education and Research investigated exactly this question.

What is notable is the direction of travel. The technology is not meant to replace the knowledge but to take over the capture — the part where people fail, because they cannot say what they know.

Four things that make knowledge transfer possible at all

From the evidence, a picture emerges of what makes knowledge retention fail in production — and what carries it.

It starts years before retirement. The KOFA study explicitly recommends planning tandems, mentoring and structured handovers so that experience is not passed on only shortly before someone leaves. Starting in the final quarter is starting too late.

It costs the experienced person as little time as possible. Any method that demands two hours of writing will not be used — not out of unwillingness, but because the time genuinely is not there. Recording what is being done anyway is the only form of capture that does not worsen the bottleneck.

It shows rather than describes. What cannot be phrased can be demonstrated. That is why video-based formats carry tacit knowledge better than prose — and why they work across language barriers where translated documents fail.

It ends up where people look. Captured knowledge that lands on a shared drive is as lost as the unwritten kind. It belongs in the knowledge base, the intranet or at the workstation itself — findable at the moment the question arises.

Put together, this leads to the same conclusion as the new ISO 9001:2026 does for process documentation: what counts is not that something was documented, but that it lands.

Frequently asked questions about knowledge transfer in production

What is experiential knowledge?

Experiential knowledge is the capability that grows out of repeated practice and goes beyond subject expertise. It shows up above all in difficult situations: experienced people resolve them quickly, inexperienced people take considerably longer. A large part of it is tacit — not fully conscious even to the person who holds it.

Why is writing processes down not enough?

Because the decisive part usually cannot be written down. Someone who has performed a task for years no longer knows which details need explaining. Documentation captures the steps, not the judgement in between.

When should a company start transferring knowledge?

Well before retirement. Labour market research recommends planning handovers structurally rather than triggering them by event. In practice that means several years of lead time for key positions, not several weeks.

Which methods demonstrably work?

Learning by doing combined with mentoring and age-mixed teams is regarded as the most effective route for tacit knowledge. The limitation is not the method but its availability: it consumes the time of experienced staff, which is scarce in a labour shortage.

Can AI preserve experiential knowledge?

Not replace it, but help capture it. Research projects such as KI_eeper pursue the approach of detecting patterns in the running work process automatically, instead of asking employees for a description they cannot supply. The decision stays with the human.

What helps with a workforce of many mother tongues?

Formats that show rather than describe, and translation in addition to the image rather than instead of it. Plain text translation of work instructions is considered only partly effective and error-prone in research.

How do you measure whether knowledge transfer is working?

By onboarding time, by the error rate in the first months, by the number of questions still routed to individual knowledge holders — and by whether documented content actually gets opened. Usage data is the most honest indicator.