This was a super successful week for the superhero cape team! Eve and I managed to put in a lot of hours, tested many things, and have a much clearer vision for our final project. The main issue we faced this week was finding the right microcontroller for our project.
Last Thursday, which feels like a year ago, we met with Benedetta who helped us troubleshoot our issues with using the BlueFruit with the Arduino Mini. Two important things we learned was that there was an issue with the BlueFruit and accelerometer we were using. We spoke a lot about how we could calculate speed with the accelerometer. The advice she gave us was to go back to testing on the Uno until we are certain about the data we are getting. This was hard personally to go a few steps backwards, but retrospectively it was all so worth it!
As recommended by Benedetta, I went back to the serial lab input labs when we graphed data from Arduino. First I mapped all three axis points in different colors, and then just limited it to the data we will actually be using, which is mainly the z-axis. This ended up being a great way to initially tests other things.
When we started to feel more comfortable with the data, Eve worked on a cape calibration programmed through p5, which turned out AWESOMELY!
Since the data aspect was good, I went to focus on starting to fabricate glove prototypes. After talking to Teresa in the Soft Lab, we spoke about the idea of making fingerless gloves, which will eliminate anticipate issues with sizing. I then whipped up a shiny, purple glove.
Thankfully, Friday was Soft Club meeting was all about big switches. It was great to talk with other ITPers who are interested in soft fabrication and soft circuitry. Plus, it was awesome to discuss and walk through ideas. I have been progressively getting more and more interested in this topic. Jason Beck is very knowledgeable about soft circuitry with his amazing work in assistive technology and very generous with his time (and basically has anything you could imagine to borrow).
Having the glove prototype helped me to visualize the rest of the project. For the past few weeks I have been designing the cape to have the gloves attached to sleeves that would be sewn in. This would mean that the microcontroller would live in the back of the cape and wires would run down to each arm. This would be kind of complicated, especially for someone with no experience with soft circuitry at this scale. Teresa and I discussed this a lot and I would just need to make sure the wires are secured, but also flexible to accommodate the users’ movements. This made me very nervous as I was uncertain I’d be able to successfully build this.
I put that aside on Saturday and worked on implementing the accelerometer data inside of a sketch made with a webgl canvas in p5.
The complicated part of this was making sure the incoming data was adding to the camera. Initially they were at the same speed, which meant each time the accelerometer would drop, or return to the resting state the camera would then jump back.
This was a great success for us and allowed us to think more about our next prototyping tests.
Saturday took a turn for me later in the day, however. I began thinking more about the cape and all of the wires and the uncertainty of not knowing which micrcontroller we would be using. I am not sure why it to took me so long to realize that by having the gloves attached to the sleeves of the cape would hinder tall people to use the cape. Then I decided to consider having a microcontroller in the actual glove. This would eliminate my concerns with the wires and allow the cape to have more movements.
On Sunday I spent most of the day exploring microcontrollers: sensorTag, RFduino, and Moteino. We reached a place where our uncertainty of microcontrollers led us to be swayed in every direction that came our way. Teresa emailed me Sunday night to tell me about a microcontroller that might be “the one”– The Adafruit Feather 32u4 BlueFruit LE. Finally, we decided to take office hours with Tom for Monday to discuss this further.
After meeting with Tom he helped us understand more the data we were receiving. He looked more into the Adafruit Feather and confirmed it would be a good choice to try out. So, we bit the bullet and ordered one with a lipo battery. In the meantime he advised us to test the accelerometer on a Leonardo, as its processor is the same to the Feather, meaning once we get the Feather it would a simple transition.
After that I spent some time trying to work with Adruino Mini + BlueFruit. This worked with the accelerometer. So I decided to solder more wires on the board to accommodate a switch and make it more compact to test with. That led to an hour of soldering that should have taken less time and ultimately did not work. For sanity’s sake, I put that aside and went back to the Uno. My goal was to not only get the accelerometer working, but also a switch in order to emulate our final project.
This was very challenging for me. The data coming in for the switch and accelerometer was fine in the serial monitor and the console in p5, however, the push button wasn’t changing the background (as I coded it to), but rather it sped up the camera. That’s when I realized the data was interfering. Finally I was able to get it to work.
The next challenge was doing that same thing on a Leonardo. Simple, right? 7 hours later…
Now I feel ready for the next steps. Prototyping this inside of a glove.
As for the cape, I worked on simulating the cape’s fun effects when the user puts it on.