March 14, 2008
Input Devices: What You Do != What Is Sensed
Together with Sean Follmer and Pat Hanrahan at Stanford I've been thinking about ways to describe and model the design space of input devices.
One important point from our discussions is that the action the user of an input device performs is often very different from what is actually sensed by the device. This difference is rarely discussed in HCI literature, yet the people who design input devices are acutely aware of it. Toy designers for example are experts in exploiting it. Since toy development is driven by cost, almost all toys will only contain the most elementary I/O components: buttons and LEDs.
Take the game "Bop It" for example: It's a variation on "Simon Says" in which you have to perform different actions ("bop it, spin it, twist it, flick it, pull it") on cue.
Yet inside Bop It, there are no potentiometers, sliders, encoders or flex sensors - everything is done with simple momentary switches and some ingenious mechanisms.
For example, in the "spin it" action the user spins a plastic gear. Inside, the gear has a recessed groove that acts as a cam. A plastic arm follows the groove while pivoting around a point at the center of the arm. When the groove moves away from the center of the gear, the arm pushes a momentary switch on a circuit board:
