Yesterday, we looked at how the Gothic cathedral architects of northern France used precedent to guide what could be built next.
But what happens when there’s no precedent?
When Antoni Gaudí was designing the Sagrada Família in Barcelona, there was no precedent for the complex geometries he wanted to build. So, he created a model. A very ingenious one: using hanging chains and sandbags to mimic the geometry and loading of the cathedral’s roof.
This physical model acted as a lead indicator, giving Gaudí insight into whether his structure would stand up. When there’s no precedent, you can’t ask “does that look right?”—because you’ve never seen it before.
The reliability of this kind of lead indicator depends on the accuracy and appropriateness of the model. Selecting or creating the right model improves with training and experience.
Engineers build models all the time. In fact, every engineering calculation is a model of the future. A structural stability calculation gives us a lead indicator about whether a structure is going to stand up. Engineers work very hard to make sure these models are as accurate as possible.
But they are still just models. The truth comes after the fact: did the building actually stand up? That’s the lag indicator. And in those rare cases where something goes wrong, this new knowledge gets fed back into better models for the future.
Thankfully, very few buildings in the UK fall down due to bad modelling. That’s because this feedback loop—between model, reality, and revised model—is quite advanced.
But what about the other areas of engineering where we don’t close the loop?
That’s a question for tomorrow.