Welcome back to Input Origins, our monthly time machine through the evolution of control.
Sitting exactly where you are, you're bending every electromagnetic field in the room. You block light. You reflect radio. You carry a hundred-odd picofarads of capacitance to ground. Congratulations! you are already an input device! You're just not plugged into anything, yet. And for almost a century, engineers have tried to plug you in. The trick, each time, was the same: pick up on the human body's manipulation of waves to generate input.
Series Editor: Ariel Amar
1920: A Tech Demo for Lenin
It started as a mistake. In 1920, Soviet physicist Lev Termen, later known as Léon Theremin, was building a precision gas-density meter, and added an audio circuit to make tiny capacitance shifts audible. He noticed his own hand, waved near the antenna, was shifting the tune of the circuit.
A cellist, he picked out Saint-Saëns' "The Swan" by ear, played it for Lenin two years later, and licensed the instrument to RCA. It hit music stores on October 14, 1929, ten days before Black Thursday, promising "Anyone who can carry a tune can play the Theremin!", but almost nobody could. Under 500 units sold.
Léon Theremin playing the Theremin
Infrared: From the Cockpit to the Couch
Infrared light is electromagnetic radiation just below the visible spectrum, felt as heat and not seen. William Herschel discovered it in 1800, measuring temperatures through a prism and noticing something warm lurking past the red end.
A century passed before engineers found uses for it. The first use as human input came in 1967, when NASA set out "to make it possible for man to control a machine by eye." The Honeywell Oculometer used infrared to track a pilot's pupil; its military sibling, the VTAS, let pilots guide missiles by looking at the target. Neither was ever adopted.
Infrared found a quieter second act in 1972: the PLATO IV terminal used a grid of IR beams to build one of the first touchscreens (Input Origins #13). But IR as input went mainstream in the living room. The Philips RC4 remote (~1977) beat its ultrasonic competition (Input Origins #1) not on engineering, but because people's keys, coins, and dogs kept accidentally changing the channel.
The Philips RC4F the first IR remote control - Marcel's TV Museum
Radar: From Bombers to the Google Pixel
Radar works by transmitting a signal and measuring what comes back: velocity, range, and motion, all recoverable from the echo. Cameras and IR sensors need a clear path, a lit scene, line of sight. Radar can work through walls and dark rooms, and doesn't carry the privacy concerns cameras do.
By the mid-1990s, miniaturization made radar components small enough that engineers started experimenting with detecting human motion. The first instance took place at MIT's Media Lab: the "Magic Carpet" (1997) used Doppler radar sensors to read a performer's gestures and drive a live audio installation.
In 2016, Israeli company Vayyar repurposed its breast-cancer-detection radar as the Walabot, a consumer device that tracked body movement through drywall. Then Google: Project Soli shrank a full 60 GHz radar transceiver onto a 6×5mm chip and installed it in the Pixel 4 to read the micro-gesture of a single finger.
The Google Soli, a 6×5mm radar chip reading micro gestures.
A century of engineering, four different bands of the electromagnetic spectrum, and the target was often the same: the human hand in free space. None of them fully cracked it, just yet.
