title: "Why structural design is an optimization problem" date: "2026-05-20" excerpt: "The slowest part of structural engineering isn't drawing — it's the search for an efficient, compliant system. That's a problem machines are good at."
Ask any structural engineer where their time goes and you'll rarely hear "drawing." The real cost is iteration: propose a set of sections, run the analysis, check the capacities, find the overstressed members, adjust, and repeat. Each loop is hours.
A vast discrete search space
A lateral system is assembled from a catalogue of standard sections. The number of valid combinations is enormous, and most of them are either non-compliant or wasteful. Finding the efficient, code-compliant corner of that space by hand is slow precisely because the space is so large.
Framing it as optimization
Once you see sizing as a search over a discrete space — with code checks as constraints and material efficiency as the objective — the path forward is clear. This is an optimization problem, and optimization is something we know how to automate.
In the next post we introduce Structural Design, the first product built on this idea.