Insect flight mechanics
Why Insects Fly in Ways That Shouldn't Work
For decades, engineers calculated that bumblebees couldn't possibly fly — and the bumblebees, entirely unbothered, kept flying anyway.
The Idea
The old aerodynamics textbooks had a problem with insects: the math didn't add up. Classical steady-state aerodynamics, the framework built to understand how aircraft wings generate lift, predicted that insect wings were too small and beat too slowly to keep a body aloft. The insects, of course, were not consulting the textbooks. The resolution came when researchers stopped treating insect wings like miniature airplane wings and started studying what the wings actually do. Insects don't generate lift the way a fixed wing does — through smooth, continuous airflow over a curved surface. Instead, they exploit unsteady aerodynamics, a messier, more dynamic regime where the interesting physics lives in vortices, timing, and controlled instability. Three mechanisms do most of the work. The first is the leading-edge vortex: as a wing sweeps forward, it peels a spinning column of low-pressure air off its front edge and rides it like a surfer on a wave — for just long enough before the vortex destabilises and sheds. The second is rotational lift, generated in the brief moment when the wing flips between strokes. The third is wake capture, where the wing swings back through the turbulent air its previous stroke left behind, reclaiming energy that would otherwise be wasted. What makes this remarkable is that insects are not just tolerating aerodynamic complexity — they are actively harvesting it. Chaos, for a dragonfly, is not a problem to be engineered away. It's the engine.
In the World
In the late 1990s, Michael Dickinson and his colleagues at the University of Washington built a machine they called Robofly — a robotic fruit fly wing scaled up to the size of a human forearm and submerged in mineral oil, which is thick enough to mimic the aerodynamic conditions a real fly experiences at its tiny scale. It was an ungainly contraption, but it let them film and measure what actually happens during each wing stroke in slow, viscous detail. What they found upended assumptions. When they isolated the rotational phase — the split second when the wing reverses — and varied its timing, lift changed dramatically. A wing that rotated slightly ahead of the stroke reversal generated substantially more lift than one that rotated behind it. The fly wasn't just flapping; it was precisely timing a flip. And that timing, they showed, is actively controlled by a cluster of tiny muscles that make micro-adjustments at up to 200 beats per second. Dickinson's lab later worked with high-speed cameras to catch free-flying flies responding to visual disturbances — and found they could alter wing stroke timing within a single beat. The control system is operating at timescales close to the physical limits of muscle tissue. The dragonfly, meanwhile, turns out to be even more sophisticated. It flies with four independently controlled wings and can set them beating slightly out of phase to generate constructive interference between their wakes — essentially using one pair of wings to pre-condition the air for the other pair. Dragonflies catch over ninety percent of the prey they pursue, a success rate no other aerial predator comes close to matching.
Why It Matters
There's a version of this story that's just impressive biology — and it is, undeniably, impressive. But the deeper implication is about the relationship between complexity and control. Engineers designing drones and micro air vehicles spent years trying to eliminate unsteady aerodynamics, treating turbulence and vortex shedding as noise to be suppressed. Insect flight mechanics suggests the opposite instinct might be more powerful: instead of fighting the complexity of fluid dynamics at small scales, design systems that are tuned to exploit it. That shift in framing — from 'chaos as problem' to 'chaos as resource' — turns up in other places too. Neuroscience is finding that certain kinds of neural noise improve signal detection. Ecological thinking has long understood that disturbance regimes can be integral to ecosystem health. The insect wing is a physical demonstration of a principle that keeps recurring: that robust, adaptive systems are often not the ones that eliminate variability, but the ones that have learned to work with it. Next time something is working in a way that shouldn't, it might be worth asking not how it survives the chaos — but whether the chaos is precisely the point.
A Question to Ponder
Where else in your life or work might you be spending energy suppressing complexity that you could instead be learning to use?
Get a new one of these every morning.
Start learning with Thinkable