DESIGN INTELLIGENCEJune 26, 2026·Mary · DEPIX Design Intelligence

BMW spent ten years on a wave nobody used. Now it's gone.

For ten years, the most photogenic thing in a BMW cabin was a hand. Twirl two fingers clockwise and the stereo got louder. Jab toward the dash to take a call, swipe sideways to reject it. Gesture Control debuted on the 2015 7 Series as proof that a luxury car could read you in mid-air, and every auto-show stand since has run the same demo: a presenter's fingers conducting the dashboard like an orchestra.

At CES in January 2025, BMW confirmed it is over. The company's next interface, iDrive X — arriving first in the Neue Klasse iX3 — ships without Gesture Control at all. A decade of development, a genuine world-first, quietly retired, because BMW's own telemetry told it what every passenger already suspected: almost nobody was using it.

That is the part worth sitting with. This was not killed by a rival or a regulator. It was killed by data BMW could only collect after the feature shipped to hundreds of thousands of cars. The studio loved it. The auto shows loved it. The owner waved twice, felt faintly ridiculous in front of a passenger, and never did it again.

The failure was baked into the interaction, not the silicon. Gesture Control demanded that your hand sit inside an invisible box — a zone defined, roughly, by one line rising from the shifter and another from the top of the air vents. Too high, too low, too quick, and nothing happened. As BMW Blog summed it up in November 2025, the system "never became easier, faster, or more natural than the controls already in place." A volume knob asks for one thing: turn it. The wave asked you to learn a vocabulary, perform it precisely, and accept that it might ignore you — all to do something a button did instantly. BMW even rebranded the feature "Natural Interaction" in 2021. Renaming a thing does not make it natural.

None of this needed a decade and a production fleet to discover. The real question — does waving beat the control it replaces? — is answerable in the concept phase, before a single time-of-flight sensor is specified, if you test the decision instead of the demo. The demo always works: a trained presenter, a quiet room, a memorised gesture. The decision is whether a distracted driver in traffic, with someone in the passenger seat, will reach for the wave over the knob. Those are different questions, and confusing them is exactly how a feature ships beautiful and dies unused.

This is the gap design intelligence is built to close. A parallel design team that can simulate the decision — model the interaction, pressure-test it against the physical control it competes with, surface early that the wave loses on speed, reliability and plain dignity — catches the gesture problem in week one, not year ten. The point was never a prettier rendering of a hand in mid-air. It was knowing, early, that the hand loses.

BMW is now betting the cabin on voice — an AI assistant it says is good enough that you never lift your hands from the wheel. That may well be the right call. But it is the same wager gesture made: that a new modality beats the established control. The only thing that has changed is whether you find that out before you commit ten years, or after.

Gesture Control was not a stupid idea. It was an untested decision dressed up as an inevitability — and today's cabins are full of those: screens that bury the hazard lights, capacitive sliders that need a glance to hit, ambient theatrics that photograph better than they work. The studios that win the next decade will be the ones that can tell a demo from a decision before the tooling is cut.