To get started on my final project, I first had to decide on a theme. My brother is getting married in September and I wanted to make his wedding gift from scratch, so I decided to make him and my soon to be sister in a law a clock for their future home. I knew I wanted to create this clock using embroidery framed in an embroidery hoop and sewable LEDs. I first had to decide what size hoop to use. I landed on using a 14 inch hoop because after research clock mechanisms, I felt the 14 inch hoop allowed for enough room between the edge of the second hand and the edge of the clock, while still keeping the clock from being too big. I knew I wanted to embroider a night scene that was meaningful to them and include the LEDs as stars in the background, so I first set out to find the perfect reference image. I decided on using the image you see below, which was taken on a camping trip Mark and Emily went on a few years ago to the Sand Dunes in Namibia. I had fabric I knew I wanted to use for this, but it was a neutral white-ish color, so I had to get black fabric dye to make sure I had the nighttime feel to my scene.
First, I died my fabric. I had to die it a few times over to get the right black. Then, I added the fabric to the embroidery hoop. I then traced the reference image using tracing paper and transferred the outlines of the objects I wanted to include over to the fabric via transfer paper.
Once the tracing was done, I was able to embroiderer my scene! This took a few days to complete as I wanted to make sure it looked as accurate and professional as possible. The hole you see in my fabric below is where the clock mechanism will be located. Once I finished the embroidery, I began planning my electronic implementation. I decided against stitching the electronics directly into my fabric to avoid an unfinished feel. So instead, I sized and cut out two pieces of felt to back my fabric to attach my electronics to.
I decided on using the Lilypad Lilymini to power and control my LEDs. The Lilymini is small, can control a few dozen LEDs, features a spot for a small CR battery, has an on/off switch and has a port to upload new code (which surprisingly not all Lilypad products have). I bought the Lilymini in a snap out kit, which included 4 sewable LEDs, a button to change the LED setting, and a light sensor. Using these products, I decided to code the light sensor to control the LEDs. When there is light in the room, the LEDs are off, and when the room is dark, the LEDs turn on. I also decided to code my LEDs flicker to make them look more accurately like stars. I also decided using the provided button was not necessary as my LEDs only have an on/off setting controlled by the light sensor. It took me a bit to get this code working on the board as the board is a bit finicky and would shut down and require a reset often when I was trying to load the code on. Once I got the code working correctly, I was able to start planning the placing of my LEDs.
Next, I had to decide on where I would put my LED stars in the sky. I decided against recreating the sky from the camping night and instead on recreating the night sky from the night and place when Mark and Emily first met: September 27, 2012, Lusaka, Zambia. I wasn't 100% sure what direction the image of the camping scene was facing, but I took my best educated guess. I looked at the meta data of the photo and saw it was taken at around 5am. You can see light appearing on the rocks behind the car, so I assumed this may be from the sun rising, which means the camera was pointed away from the rising sun, so I guessed we are looking west. I used this website to recreate the night sky on their first night, and looked at the sky with west on the bottom. I picked four constellations that you would be able to see when looking west and drew these out on my backing sheet. I placed 17 LEDs on the spots where the most visible stars shine in each constellation and hot glued each LED in place. Each LED is white, except for one orange LED where the star Antares is located (as this star appears orange/red in the night sky).
I then moved onto soldering these LEDs to wires to connect them all to the microcontroller. I am currently still in this process. Stay tuned! :)
For this assignment, I worked with Tyler Milligan and Michael Nicolaou. First, we started by saying that both microbits would be healthy (shown by a happy face). We then added that if you press the A button on either microbit, that one becomes infected (shown by a sad face). We used a probability statement that used the signal strength of the microbits to determine how likely the second microbit would be infected if one became infected. The closer the microbits were to each other, the more likely (and quickly) the second one would be infected. The further the microbits were from each other, the less likely (and more slowly) the second microbit would become infected. Our probability statement also said that if the microbits were within signal range of eachother, even if they were at the furthest point away from eachother, the second microbit would still eventually (and very slowly) become infected. We also included that if you press the B button on either microbit, it will become uninfected, just for debugging purposes.
We were running into a strange issue where the first microbit got infected, but when the second one did it would reverse the infection in the first one. We also had a few issues getting the signal strength to work correctly, so we had to fix and re-upload our code many times.
But, we finally got it to work correctly!
First, I came up with the concept to my project. With inspiration for Billy Porter's hat from the Grammy's, I decided to make a hat that covered one's face when they pressed one button, and uncovered their face when they pressed another.
My next step was to set up the code for my servo motor, what I would use to move my parts, and the two buttons that turned the motor each way.
Next, I tried to figure out what I could attach to the motor to hold the piece of fabric that would sometimes be covering ones face. In attempt to do this, I attached a cardboard square to my servo motor.
I then tried to find a hat that would work my concept. I couldn't find a hat that was shaped correctly to easily attach a moving sheet to the top. while I was at Goodwill searching and came across a lampshade and had the bright idea (ha, get it), to cut a hole out of the lampshade and use the servo to have a piece of fabric cover this hole when one of the buttons are pressed. I also purchased a cheap straw hat in attempt to use these in conjunction with each other so it would be easier for a user to wear the lampshade as a hat.
After a lot of attempts and a lot of time lost trying to get a piece of fabric to cover the hole in the lampshade, I decided to scratch the idea and try just working with the straw hat. This unfortunately did not work either.
I then attempted to use one of my own bucket hats and attach the servo to the top of it, in hopes that a hat that fit better on the head would help solve the problem of the servo motors balance, stability and ability to turn the cardboard (and later a hot plate was added on to give the servo a wider range of turning). I attempted to put on the fabric onto the hot plate but was met with the problem again that the fabric would stick to the hat as it was attempting to move the fabric away from the face, so it wasn't successful still in moving the fabric.
After re-reading the description of this assignment, I was unsure if I had a shape changing wearable with this face covering hat as the piece of fabric was moving, but the hat itself wasn't changing shape. I decided to try to make a Hi! and Bye sign that would raise up or down depending on which button you pressed. However, this attempt also was a failure and only moved a few centimeters instead of completely upright or flat against the hat like I was going for.
After many hours lost trying and failing to get this face covering hat to work, I decided to take a step back and simplify my piece. I added a cardboard square on top of the one connected to the servo, that turns from Hi! to Bye when you press the button. So you can tell those around you hello or goodbye with a quick press of a button!
The first step of creating a wearable to support navigation without visuals was to decide on my idea. I decided on using radio communication between two microcontrollers to create a controller operated navigation system. One person is holding a microbit and the other person is wearing a microbit bracelet that is connected to two vibrating motors, one on the right side of the users wrist and one on the left side. When the button A is pressed on the first microcontroller, it sends a signal to the second microcontroller to display "left" and vibrator the left motor. The same happens for the right side when button B is pressed. Writing the code to get the microbits to radio communicate was quite difficult and took a lot of time but I was able to finally figure out how to get the code to do as I planned with the help of Mary West.
The second step was to start building the circuit. The last wearable project I did with vibrating motors did not work. Thanks to the knowledge of Mary West, I now realize the reason I couldn't get the motors to work the first time was because I needed a transistor between my microbit and the motor as the motor was short circuiting the microbit and will not work without a transistor. I found this out on Thursday of this week and ordered the parts I needed which unfortunately I was not able to get until Sunday. However, once I got ahold of the transistors I was able to build my circuit and get the motors to vibrate on command.
After I got the circuit to work as I wanted in the breadboard, I worked on converting this circuit system to a wearble. Since I was only able to work on this on Sunday night and I do not have access to wire strippers or a soldering gun, I had to use electrical tape to hold my connections together. Since these connections aren't very secure being held together by only tape, my motors are a bit finicky. Other than the unstable connections, my wearable navigation guide worked as planned!
To begin my sensory glove, I first decided on using a flex sensor to turn on LEDs that would be attached to the fingers of the gloves. Having difficulties in my last wearable project attempt, I decided to break down the process by testing each part of my glove one piece at a time. I first connected a flex sensor to my micro:bit and learned how flex sensor was sending data to the micro:bit by printing out the data the flex sensor was giving the micro:bit in a display scroll across the screen of the micro-controller. This part of the process gave me the most trouble, I had a very hard time finding code to direct me in how to get the sensor to tell the micro:bit what to do. I realized I had to graduate to writing in Python rather than the block code micro:bit provides. This was especially difficult because I have never writing in python before, so I ran into a lot of syntax errors in my code that didn't allow my micro:bit to work correctly that took me a long time to debug. Once I got the code working learned what numbers I was working with when the sensor was bent vs straight, I was able to test the control of my sensor by telling the micro:bit to change from a square to a heart when the sensor was bent. After I got this working, I was able to add LEDs to this circuit and code the sensor to also turn on the LEDs at the same time the square would change to a heart. I originally wanted a different colored LED for each finger, but I realized the micro:bit didn't have enough power to control the blue and green LEDs so I stuck with just using red, as that is the color that requires the lowest amount of energy.
After I got the basic coding and circuit completed, I moved to making my circuit the correct shape I needed to apply it to a glove. I soldered each pin of each LED so each ground and power were connected to require as little connections needed as possible. I made it so I only needed three alligator clips to connect to P0 (to the power of the sensor), P1 (to the power of the LEDs) and ground that connected to both the ground of the sensors and the LEDs. After soldering each piece, I tested the circuit again in the breadboard to make sure everything still worked. I ran into more syntax errors with my code here that caused my circuit to not work and caused me to think I messed up my soldering, but luckily I was able to debug and only the code was a problem!
I think secured my circuit to my glove: one LED per finger, and the flex sensor secured to the first finger, triggering the LEDs to turn on and the display of the micro:bit to change from a square to a heart when the user's hand is bent.
I am really happy I was able to make my glove work. I know this isn't the prettiest wearable and the circuitry and adhesion onto the glove isn't the most secure, but this is the first time I was able to control an input and an output through a micro:bit successfully and I am very happy I was finally able to figure it out! Baby steps!:)
Welcome to my progress blog for my Wearable Technology course at the University of Colorado, Boulder. Enjoy watching my process as my ideas become a reality.