After last week’s playtesting, my main goal for the next week was to more accurately map the raw data coming in from Arduino to p5, finalize the cape fabrication, and solder all the hardware to fit inside the glove. Eve and I also discussed making the narrative more clear in the game.
We received feedback that the user was confused if they were still flying around the planet and wanted to see that feedback shown in the final flying scene. We decided that we would keep the same narrative but move the mission scene to before “Flight School”. That way the user would be clear on why they were learning to fly. We also discussed having the user flying through different levels of space; the higher you would fly, the weirder the images would be that you fly through.
As you can see from the video below, the speed in the game isn’t as accurate to the movement of the accelerometer. The good parts are that the BLE is functioning great! The data is coming in through the terminal. I spent Saturday working on the math in p5 to better map the data coming in. What I was noticing is that in the Serial monitor, data for the button and z-axis of the accelerometer was coming in perfectly. However, when I ran it through the terminal, the data seemed to dump out, which meant that the Feather BLE must be collecting data as it’s running and then once the terminal opens, the data would just pour out.
I reached out to Jingwen, the second year, who has a lot of experience and interest in wearables. She helped me understand how to run the BLE through the terminal using Node.js and mostly helped me understand the code Tom had helped us with in office hours. She recommended talking to Abishek, who she worked with in a wearable project last year, and feels very comfortable with Node. I then reached out to him to explain the issue we were having with the data streaming in too quickly. He was very generous with his time and helped tweak the node code a bit. We added this if-statement–since the data was streaming in the serial monitor fine with the button and z-axis, we wanted to make a line that said that only if the data is 5 characters, then send the data through the terminal. This ended up helping a lot and also slowed down the data coming into p5.
Below are two videos that showing the Feather BLE working fairly well with the p5 sketch.
On Monday, I finally received the Adafruit order I was waiting for: 10mm white diffused LEDs, conductive velcro, and a strand of EL wire. I spent the rest of the day constructing the cape.
The first thing I did was draw out the circuit with the LEDs. I knew I wanted them in parallel and to have them all connected to one battery.
So, then I got to work. I cut small pieces of stranded wire and stripped each end. Then I rolled the legs of the LEDs and soldered the stranded wire to the positive and negative sides. Once I had 2 strands of 4 lights, I wanted to see if I can add one more to each, but was nervous about using a 3V battery. I spoke to Dhruv, who was working in the soft lab and he said that using a higher voltage battery (like a 9V) would work, but should be tested for resistance. We connected everything to a potentiometer and moved it until it reached the brightness that I liked. Then we measured that with a multimeter, and it came out to 220 ohms. Then I soldered a resistor on the strand coming from the positive wire of the battery that would go up to the conductive velcro.
I was still connecting the two LED strands with alligator clips. I was ready to solder a positive and ground wire to each LED strand to connect the entire circuit.
The next thing I had to do was make space for each LED. Since I had already sewed the gold and blue materials together, I lined the LED strands on the edge and marked where each LED would be. Then I used the seam ripper from the soft lab and pulled out the thread where the LED would go. After it was time to test that the LEDs would fit. I was so surprised that this actually worked! It was time to check that the circuit still worked. I flipped the cape from inside out and pinned it to the mannequin. I used the alligator clips I borrowed from Zoe to connect the stranded wire to the conductive velcro. And yay, the circuit still worked! At this point, it was about midnight on Monday and I was deliriously tired, so I decided to head home.
On Tuesday when I returned, I thought I would take a break from the cape and work on the glove. I have been keeping the accelerometer and Feather microcontroller connected to the breadboard. My first, but painful next step was to downsize everything, as it would need to fit inside the glove. I had purchased quarter-sized proto-boards from Adafruit. I took the accelerometer and Feather off the breadboard just to test the size on the protoboard before soldering. I went to talk with Justine Peake, who I also had made glove in his pcomp final last year. He helped me to figure out the most effective placement for the parts. Since we were using a button to test what would eventually be a soft switch inside the glove, I thought that I should test the button with wires first. I also knew that I would need to solder a wire or headers to the Feather. I went back to test the button with the same set up on the breadboard. When I went to run the data in the serial monitor, something very strange happened and the numbers coming in were completely different than the had been for the past 3 weeks. I was really worried about this, as we had mapped all of the raw data to p5 in order to create the speed simulation. I knew that we shouldn’t expect the same results always to come in on an accelerometer, but I thought it was strange that the resting number had been around 610 and then suddenly it was 47. I thought maybe I just need to restart Arduino, my computer, etc, so I did, but the same results came in. I ran it in the terminal and the BLE turned on (blinking blue light) and would say that it was connected, but data would not stream in. I knew that I was beginning to panic, so I decided that I would just test the button. I ended up sewing stranded wire into the conductive fabric inside of the soft switch in the glove using conductive thread. If anything for the final presentation, I could just put the entire breadboard inside of the glove.
Going back to the cape: After the strands were connected, I used conductive ribbon and sewed the stranded wire using conductive thread into the ribbon. Then I sewed the conductive velcro to the conductive ribbon.
After I wanted to make a small enclosure for the 9V battery. So I made a little sleeping bag using scrap fabric from the soft lab. Then I sewed the entire bed into the top of the cape. This was a last minute decision, and although it worked, I wish I would have included it in the initial design. I think then it would have been more secured inside the lining of the cape.