After scrapping those 300 components, one thing became very clear: we needed a better way to measure.
The machines we had bought were solid and fully capable of producing high-quality timber components. But their degree scales were designed for woodworking where whole or half degrees are often enough—not for a geodesic structure where the difference between 6.8 and 7.0 degrees really matters.
Everything was clear in the CAD model. One bevel might be 6.8 degrees, another 7.2 and a third 4.1. In the workshop, those same values suddenly depended on how precisely we could read a small mechanical scale.
That was not enough.
So we began looking for precision instruments of our own to help translate the accuracy of the model into real machine settings.
Among the most important were small digital angle gauges that could be attached directly to saw blades and other machine parts. First, we established an accurate reference plane and zeroed the instrument. Then we moved the gauge to the blade and read its actual angle directly from the display.
Suddenly the setting was no longer “about seven degrees”.
It was 6.8.
But what mattered was not merely making one perfect bevel. We had to make the same bevel again and again: run a batch at 6.8 degrees, change to 7.2, move on to 4.1—and later return to 6.8, confident that the result would be the same.
We therefore made double-checking a principle.
We never relied on a machine’s own scale alone. We verified angles with our digital instruments, checked reference surfaces and measured settings again before running a batch. We acquired different digital tools for different operations, allowing us to check saw and milling angles independently of each machine’s display.
It was not that we distrusted the machines. We had learned what one small error could do.
Measurement became part of the craft, not something we did afterwards.
Gradually, our manual workshop began to follow a digital logic: measure, verify, make, inspect and repeat.
This was also when we began to understand the difference between precision and repeatability. Making one component correctly is not enough. To build a dome with hundreds of components, you must be able to make the same component correctly a hundred times.
The precision was already there in the mathematics and the CAD model.
Now we were learning how to carry it all the way into the timber.
6.8 degrees.
Not approximately.
6.8.
During the summer of 2026, we selected three pilot partners from almost 100 applications from Sweden and abroad. The pilot year has now begun with installations in Hemavan, Åre and Borgafjäll.
Under sommaren 2026 valde vi ut tre pilotkunder bland närmare 100 intresseanmälningar från Sverige och andra länder. Pilotåret har nu inletts med installationer i Hemavan, Åre och Borgafjäll.
Tillsammans med våra pilotkunder kommer vi under det kommande året att vidareutveckla varje del av Scandinavian Domes – från materialval och hantverk till logistik, installation och långsiktig hållbarhet – inför den bredare marknadslanseringen 2027.