Michael Schubart's Donald Is a Raspberry Pi-Powered Simon Air-Solving Robot Cheat
Detecting the toy's lights as they come on and playing back patterns perfectly, Donald can beat Simon at its own game.
Maker Michael Schubart has created a Raspberry Pi-powered robot, dubbed Donald, which exists for one reason and one reason only: to cheat at the game of Simon Air.
Released by Hasbro in 2016 as a successor to the classic Simon electronic memory game, invented by Ralph Baer, Howard Morrison, and Lenny Cope and launched in 1978, which offers a twist: While the idea behind the game is still to memorize an ever-growing pattern of colors and sounds, you no longer have to push a physical button — instead, simply waving your hand near the device's motion sensors to indicate your selection.
Schubart's desire to beat the device at its own game led to the development of Donald, a semicircular robot featuring four servo motors each driving a tiny simulacrum of a human hand. The Simon Air is positioned in front of the robot, and the hands move to mimic its patterns — far faster and with greater accuracy than even the best human player.
The secret behind Donald's success at the game are light-dependent resistors (LDRs), positioned behind each light-up game segment. As the game flashes its lights at the start of the game, a program running on a Raspberry Pi single-board computer monitors the maximum and minimum light intensity on each light sensor — using that to calibrate its detection of the segments' illumination during the game proper. Once the game starts, the program records every piece of the pattern perfectly — and plays it back via the tiny servo-hands.
For those looking to set their own computer-aided high scores on Simon Air, Schubart has published the source code for the project to GitHub under an unspecified open-source license — though you will need to work out the wiring and bill of materials from pictures of Donald's inner workings.
