After the box was fabricated, the greatest challenge was figuring out how to mount the motors, which involved a significant amount of planning.
We knew we wanted the motors to be attached on the back wall of the box, which meant needing to cut holes in both the back acrylic and the box itself. We decided the easiest way to do this was to laser cut precise holes in the acrylic for the motors to sit and then cut the back of the box a little more haphazardly since it was more difficult to cut the wood.
We velcroed the back panel of acrylic to the wood so that we could easily take it out to adjust the motors. We hot glued the motors to the acrylic, but ended up velcroing a few of them in the end since it became apparent they would need adjustment or maintenance quite frequently.
The last major fabrication hurdle was mounting the bears to the motors. I’m sure there are more elegant ways to do what we did, but after much trial and error, we decided to use small wooden dowels in the shape of a cross that fit into the bears. This provided a stiff armature which to attach the the bears to the servo attachments with none other than the infamous hot glue.
We also played a lot with the positioning of the bear on the motor since the servos are particularly strong. If the attachment was too low, the bears would slump over when they were turned on.
When we put the whole thing together, we found that our soft switch, which was our very first version, was acting a little temperamental. We had to back up a few steps and quickly realized that the foam in between the conductive fabric was slipping and allowing the circuit to close indefinitely. We remedied this by trimming the conductive fabric and dabbing a bit of hot glue on the foam so it would hold in place. This worked brilliantly.
At this point, our electrical system was working very well — lights, camera, sound, motors — which was a delightful achievement. Dare I say we had fun??
Unexpected Time Consuming Tasks
1. Camera placement: We played with this A LOT. We wanted to capture the face of the user as well as some of the box.
2. Bear placement on the Servos
3. Light Color: We wanted one that was flattering
1. Add something to direct the user’s eyes towards the camera as it currently looks like people are looking down.
2. Adjust the code for when the picture is taken — it is being triggered 5 seconds after initially closing, but then takes photos instantly afterwards. It would be ideal if it would wait 5 seconds everytime so that it captures the reaction of the user.
3. Have the lights change color constantly and perhaps trigger the camera when the lights turn white.
Presenting in Class
Things we would like to change
1. Camera timing: This means playing with the code.
2. Projector quality: Right now, the photo looks very grainy when it’s projected, even though it’s quite clear on the screen.
3. Push the images to a flickr account, which we more or less already have set up with IFTT
After class, we went to move the project to a different spot on the floor to document it, but ran into trouble with the motors. When we plugged it in, all of the servos were dead, despite working literally 5 minutes before. We checked the power supply with a known functioning motor and the power and circuitry seemed to be ok.
Our best guess is that the motors were connected to the desktop power supply when we turned it on and an excess amount of power was being delivered and we surged the system.
This is what I will call Learning the Hard Way, which has been very thematic of pcomp in general.
We ordered a 10-pack of servos immediately thereafter and have plans to replace them on Tuesday.
At that point, we will take more thorough documentation of the final product.