My initial idea was to use string to put some object into tension and cause motion, but that was really difficult to control with a dc motor (though it probably would have been fine with a stepper motor). I then thought about how I might be able to flip a piece of paper with different colors on each side to convey a change in emotion via a shift in color. When I started to experiment with this I realized that the speed of the motor made it difficult to discern between color states, but that led me to stumble upon the form-changing aspects of this project.

Getting this piece to perform the way I wanted it to was incredibly difficult. There were a large number of variables, things were imprecise, and debugging was a major pain. In the end, I was able to elicit a very specific behavior out of the machine, but I think earlier versions filmed much better, though the final product looks quite nice in person and is much cleaner when not in motion than the others were.

Below is some process.

First success with a motor:

First success with a change in form:

Some iterations:

Final product:

Iterations of different components:

Files: goo.gl/qxMDw1

(uploaded to google drive because WP would not let me upload with the current filesize)