However, I didn't know how to go about making the device, and spent many hours thinking about different possibilities before a friend who has been a fabrication mentor in the past suggested that an excellent resource was right under my nose: the Etch-a-Sketch patent, filed in 1959. This was a great piece of advice. With the patent in hand, I did not need to create a novel way of moving two rods in parallel remotely; all I had to do was adapt the existing old method of doing so to my purposes.

I went through a trial-and-error process of building the device, with many, many failures along the way. Major difficulties included: tensioning the cables properly, affixing cables firmly to each other with improvised hardware, measuring cable run lengths correctly to match the bus stop installation, and tweaking the perpendicular rods so that they would lie exactly perpendicular to each other at all times and not bind in the 3d-printed junction block. I had a very hard time finding a device that could reliably cut the cable I was using, which made threading the newly cut end through the cable housing extremely laborious. Additionally, I found myself approximately 2' short of cable when all was said and done; I needed to twice modify the cable housing (involving unthreading and rethreading the cable) to adjust for the discrepancy. I tried to add a feature where a piece of paper with a drawing on it could be rolled out and ejected so the user could keep it; this was just too hard in the time I had and I abandoned that effort. Ultimately, I installed a pencil as the writing implement in the prototype device, which wasn't especially satisfactory. I substituted in a crayon which somehow was also very unsatisfactory. Many things didn't go to plan and I had to change course repeatedly throughout the fabrication process!

It would also have been nice to have had longer to observe the finished product installed in the bus stop and this would've allowed me to make further design changes in response to watching people's interactions with the machine as well as approaching and discussing it with them.

These frustrations aside, I'm glad that in the end I had a functioning (if frictiony!) prototype that I was able to install and actually observe in its intended environment. I only saw a few people actually notice that the project was even installed in the bus stop; though the project is intentionally out-of-the-way so that it doesn't interfere with people's regular use of the stop, it actually probably was too invisible. (What good is the intrigue of the box if nobody sees it!?)

Additionally, the control knobs, which are the most initially-visible part of the project because they're at eye level, frequently seemed not to have been legible as human interface devices. Metal bolts projecting out of electrical junction boxes simply do not look like things that you're supposed to touch! I would like to make more inviting controls in the next project iteration to increase people's sense of being able to interact with the device.

I believe this was an effective first prototype of this bus stop interactivity project, which was chiefly very instructive in showing the sorts of improvements that could go into making a more successful second iteration.