For my last assignment, I decided to mount a white flag of defeat to a DC motor (all in the name of good fun, of course). Thanks to James Hosken, I had a nice little L-shaped design against which to mount my actual motor:
I implemented a variation of this design, based on the 2x4 blocks of wood I had from before. As per always, I relied on the straight edge of lovely, lovely blue tape to help me make my cut on the miter saw.
Once cut, I attached the two pieces of wood using wood screws. As you’ll see below, it took a great many number of clamps to get the job done right.
I had measured my screws with my digital caliper, but sometimes measuring doesn’t freaking matter. I tested both 7/64” and 1/8” holes for the wood screws in the shop. And though 7/64” lined better with my measurements, it was the 1/8” hole that allowed me to insert my screws without disturbing the peace of the ITP shop.
I had borrowed a wooden dowel (sorry James, it’s probably ruined forever) and drilled a 1/4” hole in its base where it could engage with the shaft of the dc motor. I also drilled an 1/8” hole into its side so that I could insert a set screw to hold this nonesense in place. I also had a brass rod I wanted to to insert through the top of the dowel to hold my flag. I drilled a 1/8” hole clean to the other side (but yes, crooked, thank you for asking).
I found some spare white fabric in the Soft Lab, and attached it to my rod and dowel setup using pins (a survey of nearby peers confirmed that this would best captured the defeatism symbolism I wanted the flag to embody). Here it is in all it’s glory:
For my Two Materials assignment, I decided to keep it pretty simple. I still hadn’t made use of the blocks of wood I had prepped from my “5 Multiples” assignment, and I still wanted to have my at-home lock picking station. I decided to install my doorknobs!
The doorknobs were metal. They were pretty easy to attach to the softwood because the hard work had been done during the “5 multiples” assignment. The doorknob components fit snugly into each of the prepared holes (though I had to use a rubber mallet for the insert). There were two long machine screw for each doorknob to lock the into place with each other. Unlike so many of my other fabrication assignments, it all went rather smoothly.
There were many turns and detours between our initial prototype and the finished, wearable-playable-musical prototype. After our initial playtesting, we had a good sense of how our MPR121 would work with our conductive thread and stitching patterns. We only had to actually make the jackets! We began by ideating on potential designs for the embroidery. We had previously identified regions of touch, to differentiate between areas that would connote casual versus intimate touch.
We used those regions to guide to where we would place embroidered areas on our jacket design. We came up with the following design to use as a guideline in our construction processes.
Next, we began construction of our jackets. We took a great deal of guidance from Idit Barak and Sandy Hsieh, both of whom had previous experience making clothing. Based on their advice, we decided to trace the shape of one of our own jackets onto test fabric as the basis of our sewing pattern. We then cut out this fabric, and then traced that shape onto paper. This process allowed us to clean up some of the curves and lines that made up the shape of our jacket, as well as ensure symmetry where necessary.
When we had a complete set of paper patterns, we did final trace and cut our selected fabric. We cut two sets of the pattern for our pair of jackets. At earlier point in our project, we had intended to each design one of the jackets. We realized it would be easier if one person did the portions of the jackets, both for the sake of uniformity and ease of construction. We split the work into front and back + sleeves.
We soon found that the embroidery took quite a bit of time. There were stitching patterns that we hoped to implement because of their textural qualities that were simply too time consuming to use throughout the piece (the shop sewing machine does not take kindly to the conductive thread). In order to assemble the jacket, we first connected the seams using fusible fabric tape and then ran the seams (with a great deal of effort and frustration) on Singer Simple sewing machine.
Alongside the manual labor of making the jackets, we also had to figure out a good amount work on our technical side. In our playtest, we had our MPR121 connected to our Arduino Uno, but for the final jackets, an Uno would be too bulky and uncomfortable. Additionally, we wanted both jackets to be able to communicate with a web server that was playing a soundscape. We examined a range of other microcontroller boards that would allow a Bluetooth connection including the Lilypad + XBee, Bluefruit EZ-Link, and a Bluefruit Feather 32u4.
We ended up testing the Bluefruit EZ-Link and the Feather with the Web Bluetooth API. One of the initial challenges of using the API was identifying the correct UUIDs associated with our boards. We finally found the correct UUID on a Hackster tutorial. Beyond that point, the sample code was pretty helpful, and soon we were able to use the UART protocol to send data from the Bluefruit and Feather back to our server. To simplify matters, we decided to move forward with two Feathers for each jacket, each communicating with an MPR121 breakout board and powered by a lithium ion polymer battery.
One the issues we encountered was that the MPR121 would periodically become unresponsive. In order to solve this, we are manually resetting the board every 8 seconds in order to keep it alive between pauses in interaction. Jiwon spent a great deal of time perfecting this process into something more robust.
Even with the programmatic side was correctly working, we found that there was some work to be done with physical troubleshooting. We had sewn the Feather and the MPR121 onto a bit of canvas to be attached to our jacket. We had sewn conductive thread into pins of each of the boards. Unfortunately, during user testing, some of these threads had inadvertently come into contact, causing nothing to work. Since that experience, we have resewn the connections, and added strategically placed bits of hot glue to insulate the most critical junctures.
The electronics of the jackets are enclosed in panels of canvas, velcro-ed shut on the the back of each jacket. Jiwon created a smaller pocket to attach onto the panel to protect our lithium ion battery. The setup makes troubleshooting bit easier, and is both comfortable for the user to wear and unnoticeable from the outside.
In the end, we had 8 separate regions of embroidery on each of the jackets, each with a 4-second long musical element attached. Jiwon had used GarageBand to create samples of audio using notes from compatible chords of music, and then clipped each sample to avoid sustained sound from singular triggers.
You can see a sample of the interaction in our video here:
In our office hours with Danny Rozin, he had suggested imagine potential interaction scenarios in which two users, wearing the jackets, encounter one another. We imagined that the jackets could enable both casual exploration between unfamiliar partners as well as a more uninhibited exploration between friends or lovers. We hope the jackets invite users to explore and play with the boundaries of comfort, in touching and being touched. The piece is also a social experiment in the sense of prompting the users whether their default boundaries of comfort would be altered under the context of art and play.
I had quite a bit of trouble getting Fuse to work correctly on my computer. Often, as soon as I selected one of the pre-made torsos, the whole program would crash. It was pretty difficult and unsatisfying to build a character in this manner. I was able to get the program to stay with me along enough to build a character that had some semblance to my 4th grade teacher.
It is not as creative as I might have been with more confidence in the program’s ability to continue to live, but it was what I was able to successful export from Fuse!
This week, I’m hoping to create an appropriate enclosure for my laser etching art from last week. A very rough sketch below:
As so often happens, I feel a bit encumbered by parts of the project that seem completed unrelated to the assignment at hand. I really want to make the whole thing work without sacrificing my entire Arduino Uno to the endeavor. I’m not entirely sure what the limits to working with a microcontroller IC might be, but this assignment seems a good a reason as any to find out.
Basically, it suggests that I can use an ATtiny85 instead of my Arduino Uno to drive my Neopixels. Perfect! It seems that the maker had been able to successfully run Neopixel’s strandtest using just the ATtiny85, and I won’t need anything much more complicated than that.
The next bit I had to figure out was how to program my ATtiny85, and luckily, the internet continues to be a reliable resource for all my how-to questions. You could buy a tiny AVR programmer for $15…
OR, you could use your Arduino as the ISP. Ding, ding, ding, we have a winner. How-to, in loving detail at high-low tech.
I wanted to be a bit price-conscious for this project, and I love the possibility of making something interesting from only the components and ICs we have for free in the ITP shop.
It took a bit to get my Arduino programmed as the ISP and then have the correct Neopixel code loaded onto my ATTiny85, but it works!
I really hate dedicating my Arduino to some single-use project for any amount of time, and I don’t want to spend the money buying additional microcontrollers when we have microcontroller chips for free in the shop. The ATTiny85 is the perfect antidote to my woes – I’m going to save soooo much money, and maybe be forced to write slightly more efficient code. I heard a lot of “just make it easier for yourself” as I was working on this, but honestly, that just sounds like, “just make it more boring for yourself.”
Now to actually program in the light patterns I want. I’ve been playing around with Neopixels for other projects, so I do have some working code to reuse here. I’m holding off on constructing the actual enclosure until I know for sure what buttons I’d want to include (and of course, what their functions would be). Playing with the ATTiny85 has been really helpful in letting me see what is possible.
What ended up happening as this assignment progressed is that as I switched to finally fabricating the box in question, I realized I had little to no time for the electronics.
To design the enclosure, I modified a Makercase box by removing the finger joints. This was one of my major missteps – I forgot to take into consideration the width of the material on all of the sides. This made construction a bit tighter than I had originally hoped and made quite a bit of my measuring and planning moot.
The plywood I had purchases was also much thicker than anything I had put into the lasercutter before, and it look a really long time. It was also my first time using wood, and I was a bit displeased with how burnt the edges looked.
In the rush of assembly, I accidentally glued the front of the box to the wrong aside of an adjoining piece.
The finished piece looks like this, and I think it might be the worst thing I’ve ever made: