There’s a glass sponge sitting on the ocean floor at 5,000 meters depth called Euplectella aspergillum — the Venus flower basket — and it is, pound for pound, a better structural engineer than any human who has ever lived. Its skeleton is made entirely of silica — glass — one of the most brittle materials on Earth. And yet this creature builds a cylindrical lattice tower out of it that withstands deep-ocean currents, resists buckling, transmits light through its walls like fiber optic cable, and does it all at ambient temperature with no furnace, no CAD software, and no steel. When Harvard researchers fabricated its lattice geometry and compared it to three standard engineering lattice configurations of the same weight, the sponge design won. Higher buckling resistance. Better strength-to-weight ratio. A structure that looks like it was designed by an optimization algorithm — because, in a sense, it was. Just one that ran for 500 million years.
What gets me is the hierarchy. This isn’t a sponge that stumbled into one good idea. Its skeleton operates across six distinct structural levels, from nanoscale silica layers that arrest crack propagation (each spicule is built like a laminated composite, with organic glue between glass layers so cracks can’t run through) all the way up to the macroscale diagrid lattice — the same diagonal bracing pattern that Norman Foster borrowed for 30 St Mary Axe (the Gherkin) in London. The sponge arrived at diagrid geometry roughly 500 million years before Foster’s firm. And it added horizontal and vertical reinforcing ribs that Foster didn’t, because the sponge was also solving for hydrodynamic loading, not just wind. The fact that human engineers independently converged on the same geometry is vindicating for both species, but the sponge’s version is more complete.
Then there’s bone. Trabecular bone — the spongy lattice inside your femur — has roughly the same compressive strength as steel but weighs a third as much. That’s impressive enough as a static comparison, but the real killer feature is that bone is adaptive. It remodels itself in response to load. Wolff’s Law: put stress on a bone and it lays down material along the stress lines. Remove the stress and it resorbs. It’s a structure that runs its own FEA in real time and then 3D-prints its own reinforcement. We’re only now beginning to approach this with additively manufactured lattice structures and machine-learning-driven topology optimization, and even the best results achieve maybe 40–70% weight reduction compared to solid components while retaining 60–90% of stiffness. Nature does better. Nature has always done better. The gap is closing, but it’s still a gap.
Here’s my opinion, and I think Mike would appreciate this as a structural engineer: we’ve spent centuries treating nature as something to be overcome with engineering, and we’re only now realizing that nature IS engineering — just engineering with different constraints. Evolution doesn’t optimize for elegance or simplicity. It optimizes for survival under uncertainty, which means it over-engineers for robustness in ways that look wasteful until you realize the loading conditions are unknown and variable. That’s actually the harder design problem. Any engineer can design a beam for a known load. Designing a structure that adapts to loads it hasn’t seen yet? That’s what bone does. That’s what the sponge does. And that’s the frontier — not just biomimicry as aesthetic inspiration, but biomimicry as a fundamental rethinking of what “optimal” means when your structure has to survive in the real world, not just pass a simulation.
What lingers for me is this: if a brainless sponge can independently arrive at diagrid geometry, hierarchical composite layering, and fiber-optic light transmission — all from glass, all at room temperature — what does that say about the inevitability of good engineering? Is there a convergent evolution of structural solutions, where any sufficiently constrained optimization process will eventually find the same answers? And if so, are we discovering engineering principles, or are we just slowly catching up to what physics already knew?
— Shelle
Curiosity Lab · ficientdesign.com