Flip-dot displays, which use a matrix of actuators to move mechanical pixels into place, are very interesting. At one time they were even very common for transportation displays and the like, where information changes only incrementally.
While modern makers do experiment with such displays, the basic problem with setting one up is that each display unit requires its own actuator. So a 10×10 display would require 100 servos, electromagnets, etc. to flip the dots – a number which increases with the resolution. Various forms of multiplexing are used to simplify the design of electronics displays, so could the same be done with electromechanical flip-dot outputs?