Welcome back to Input Origins, our monthly time machine through the evolution of how users control machines.
For our 21st edition, we’re telling the story of the mouse’s overachieving brother. The one built not for pointing at a screen, but for moving through space.
It went to space. It helped drive a rover across Mars. Yet despite its technical brilliance, it never became part of everyday computing.
The 3D Mouse.
Its story shows a hard truth about input: the best interface is not always the most capable one. It is the one that matches what most people actually need to do.
Series Editor: Ariel Amar
A vintage Hewlett-Packard Spaceball 3D controller from the 1990s.
The Missing Dimension
When you examine a physical object, say, a coffee mug, you would pick it up and turn it in your hand: tip it to see the bottom, spin it to read the back. Trying to do the same digitally, however, a normal mouse falls short: it only moves in two dimensions.
That is the problem the 3D mouse solves. This stationary device sits under your hand like a small weighted knob; you gently push, pull, twist, and tilt it, and the model on screen moves exactly as your hand does. While your regular mouse points and clicks, this one turns the object over, peers at its underside, and spins it around, all in one fluid motion.
A Space ball 3003C demo.
Built for Steering a Robot
Built for Steering a Robot By the late 1970s, engineers and mechanical designers were looking for a way to manipulate 3D objects on screen. The classic mouse couldn't do the job: This meant moving an object in six dimensions instead of two, it required six degrees of freedom, or 6DoF for short, three for moving it (left/right, up/down, forward/back) and three for rotating it (pitch, yaw, and roll).
The solution came in 1981, at DLR, Germany's aerospace research lab, Gerd Hirzinger's team needed to teleoperate robot grippers in space, so they built the first 3D mouse! A hollow plastic ball, suspended on springs, wired with force sensors on all six axes. They called it the Steuerkugel, German for control ball. Meanwhile, coincidentally, on the other side of the planet, John Hilton at the University of Sydney was independently building the same device as a research thesis project for his Master of Engineering at the University of Sydney. His primary aim was to develop a "3-D force-sensing joystick" specifically tailored for Computer-Aided Design (CAD). He'd call his 3D mouse, the Spaceball!
Robotic arm in space, manipulated by a 3D controller (source)
It Went On a Space Mission
The 3D mouse first big test came in April 1993, aboard Space Shuttle Columbia, Astronauts and ground crews used a Steuerkugel 3D Mouse to teleoperate ROTEX the first remote-controlled space robot, including grasping a free-floating object in orbit from Earth across a seven-second signal delay (!)
Four years later, NASA used a Spaceball 3D Mouse to drive the Sojourner rover across Mars. Operators at JPL wearing stereo goggles, plotted waypoints on a 3D model that the rover would then execute.
The same family of devices became the secret weapon of CAD, aerospace, automotive design, molecular modeling, and industrial robotics. But for all that, the 3D mouse never quite conquered the most ordinary terrain of all: your desk.
Robotic arm in space, manipulated by a 3D controller (source)
From NASA to GitHub
So why did it never become the mouse? Well for the obvious reason that most of us don't actually navigate three dimensions digitally.
Yet the 3D mouse never died. 3Dconnexion still ships the SpaceMouse line, and a quiet open-source revival has been happening on GitHub: DIY 6-DoF controllers built around RP2040 microcontrollers and magnetic sensors, sold as kits to mechanical-design hobbyists who want to print their own. Forty years on, the device that once flew on Columbia still has a maker base soldering replacements at home.
