The Machines Built for Problems That Happen Once

When One-Off Work Becomes a Repeatable Advantage

Every tunnel presents a different combination of rock, pressure, depth, and diameter. That should make tunneling a poor foundation for scale. Herrenknecht built its advantage from precisely that lack of repetition.

Most companies scale by eliminating variation.

They standardize the product, repeat the process, and avoid work that cannot be sold twice. Herrenknecht followed the opposite path.

Each unusual tunnel forced its engineers to solve a problem they had never encountered before, but the knowledge gained did not remain trapped inside that single machine.

It expanded what the company could credibly attempt next.

Fifty years later, Herrenknecht operates across more than 60 locations because it learned how to make experience repeatable even when the projects themselves never are.

Case Study: How Herrenknecht Made One-Off Engineering Repeatable

In 1971, Martin Herrenknecht joined the Seelisberg Tunnel project in Switzerland, where he worked alongside “Big John,” then the world’s largest tunnel boring machine.

The machine repeatedly became stuck in the rock. After four years at the tunnel face, Herrenknecht believed he could build something better.

He opened an engineering office in 1975. Two years later, he founded Herrenknecht GmbH with six employees and 25,000 German marks borrowed from his mother.

The breakthrough was not designing one perfect machine. It was learning something transferable from every difficult machine the company built.

Tunnel boring does not reward simple repetition. Every project presents a different combination of geology, diameter, depth, pressure, and access. Equipment built for one mountain may never be used in exactly the same conditions again.

Herrenknecht embraced that variation. The company developed machines spanning diameters from ten centimeters to nineteen meters, giving its engineers experience across an unusually wide range of underground conditions.

Each custom project expanded what the company could attempt next. By 1984, Herrenknecht had opened its first foreign subsidiary. Four years later, it employed 110 people.

In the late 1990s, Herrenknecht built a record-setting 14.20-meter machine for Hamburg’s Elbe Tunnel. It later supplied four machines for the Gotthard Base Tunnel. Together, they removed roughly 10.5 million cubic meters of rock beneath the Alps.

The Eurasia Tunnel required another kind of solution. Its machine bored beneath the Bosphorus at depths of 106 meters while withstanding water pressure of 11 bar.

At the Brenner Base Tunnel, that accumulated expertise helped a machine excavate a record 61 meters of hard rock in 24 hours.

Projects that appeared too specialized to scale became the source of Herrenknecht’s scale. The company could approach the next difficult tunnel with knowledge gathered from dozens of previous exceptions.

That advantage now extends across global transit projects. Eight Herrenknecht machines helped bore 13 miles of London’s Crossrail, while two machines designed specifically for chalk and flint completed HS2’s ten-mile Chiltern Tunnel.

The machines were not identical to those used beneath the Alps or the Bosphorus. The expertise behind them was cumulative.

Herrenknecht now employs thousands of people across more than 60 locations. Martin Herrenknecht still owns approximately 96% of the company, while his son Martin-Devid joined its management board in 2022.

For fifty years, the company has accepted engineering work that could not be repeated exactly. Its lasting advantage came from ensuring that the knowledge gained from each project could be.

Herrenknecht did not scale by making every problem the same. It scaled by becoming better at problems that never are.

From One-Off Work to Compounding Capability

1. Preserve the Method, Not the Finished Product

A finished project may never be useful again, but the process behind it can be. Every Pixar film begins with different characters, technical demands, and creative problems. What carries forward is the studio’s method for reviewing unfinished work and exposing weaknesses early.

When the output cannot be repeated, preserve the method that made it possible. The next team inherits a better way of working rather than a formula it cannot use.

2. Move Local Solutions Across the Organization

Improvements often remain trapped inside the team that discovered them. Toyota counters this through yokoten, the practice of sharing knowledge horizontally across the organization.

A breakthrough creates lasting value when it travels farther than the problem that produced it. At Toyota, a solution developed in one workplace can inform how other teams approach similar constraints without requiring them to rediscover it independently.

3. Let Variation Improve the System

Different routes, climates, and operating conditions expose aircraft engines to stresses that no laboratory can perfectly reproduce. Rolls-Royce uses data from in-service engines to update digital models, detect problems earlier, and improve maintenance decisions.

The engines do not need identical experiences to produce shared value. Instead of eliminating variation, Rolls-Royce built a system that learns from it.

4. Make Failure Safe to Examine

When people expect punishment, they simplify what happened, protect themselves, and leave out the details future teams need. Etsy’s blameless postmortems are designed to reconstruct technical failures without searching for someone to fault.

Engineers can examine the information and pressures surrounding a decision rather than judging it with hindsight. A mistake becomes expensive twice when fear prevents the organization from learning from it. That approach helps Etsy turn isolated incidents into knowledge the wider company can use.

From strategy to legacy

Herrenknecht’s advantage did not come from repeating the same machine. It came from carrying knowledge from one unprecedented challenge into the next.

That distinction matters. Standardization is not the only path to scale. Organizations can also grow by becoming better at capturing what difficult work teaches them.

The project may happen once. The capability it creates should not.

What knowledge from your hardest work is the next team still being forced to rediscover?

- Legacy Beyond Profits