Introduction: Making a Web-programmable, Weight-sensitive Backpack
I've been traveling a lot with my backpack, sometimes putting too much in it, which is bad for my back. Taking a cue from this weight-sensing tote on Instructables, I decided to build one using schemer.
Schemer is a tiny programmable button that helps you easily make interactive art and craft. Instead of using wires or bluetooth, you program it by holding it in front of a computer screen. You only need your web browser. No wires, and no extra hardware or software to install.
It's an experiment and part of my Ph.D. research on ambient programming and wearable computation. Check out page 2 for a brief discussion of why I think simplifying the construction and programming of wearable electronics is an important goal.
With the schemer web interface on my iPhone, I can quickly calibrate it for my needs, or make it do random flashing patterns anytime I want, without breaking out a USB cable or anything.
The 3 volt felt battery holder is eye-catching too, no?
Step 1: What This Means for the Future of Etextiles (aka What's the Point?)
There are two major reasons I decided to create this type of wearable electronic button.
The first:
It hit me some time ago that the traditional ways of hooking up electronics are a big roadblock to broader acceptance of etextiles: for educators, for mass manufacturers, for DIYers, for artists, and so on.
On a circuit board, you can route all the wires carefully so they don't touch. This is easier to do on a hard, fixed, 2D circuit board where you have the luxury of routing wires under and above the circuit board.
However, in the case of sewing to fabric, the thread is typically stitched to both sides of fabric, making this kind of routing more difficult. Clothing is inherently "more" 3D than a printed circuit board: it has folds, creases, and dynamic spatial relationships (i.e. the whole fabric moves so your wiring has to account for this).
These factors add up, therefore sewing with conductive thread becomes a mess really quickly. We need to rethink how electronics and fabric may better coexist, and I think simplifying and reducing the number of connections is a good way to start.
These factors add up, therefore sewing with conductive thread becomes a mess really quickly. We need to rethink how electronics and fabric may better coexist, and I think simplifying and reducing the number of connections is a good way to start.
The second:
Programmable (wearable) artifacts still require a relatively beefy computer. By beefy, I'm referring to the bulk of all the things you need to reprogram anything in the field. I'm talking about the need for a computer with hardware ports, software, keyboard, etc. and not processor specs). Typically, I can't reprogram it from my PDA or phone.
I imagine a future where if I have a wearable electronic item and I need to change its behavior, I'd want to take out my phone, poke at the screen, and send the instructions to my shirt or pendant.
Currently, the typical way would be to break out the laptop (nettop), sit down, fire up an editor or development environment, type, compile, find a USB cable, download. Hopefully you brought all that, and you've already installed all the software and drivers you need.
I imagine a future where if I have a wearable electronic item and I need to change its behavior, I'd want to take out my phone, poke at the screen, and send the instructions to my shirt or pendant.
Currently, the typical way would be to break out the laptop (nettop), sit down, fire up an editor or development environment, type, compile, find a USB cable, download. Hopefully you brought all that, and you've already installed all the software and drivers you need.
Wearables give us a tremendous opportunity to rethink this very premise. Why can't we program directly from our phones for instance? This would be awesome especially for quick changes. Could we download a program embedded in a painting. Perhaps, by sliding my pendant in front of a painting, I could read in a program to dynamically change my pendant's colors? If I was a museum curator, could I make my visitors' light-up shirts behave differently depending on the exhibit they are in front of? What would that scenario or system look like? How would that be possible?
I feel that web-programmable wearable buttons (like schemer) are a good - though tiny - step in enabling a more pedestrian approach to etextiles. I used the word pedestrian, not to mean boring or unimaginative, but to mean commonplace and straightforward - think cellphones vs. unix mainframes.
Step 2: Ingredients, Preparation and Layout
schemer - Aniomagic's web programmable button
lightboards - smart lights that fade in out, also from Aniomagic
force sensitive resistor - from SparkFun
custom sensor board - Aniomagic
felt battery holder - Aniomagic
large 3V battery - any place really
conductive thread - Lessemf, SparkFun, Aniomagic
I wanted to have have four different lights to correspond to the weight of my backpack: green, white, blue, and red. I chose to use three lightboards and a schemer, since schemer also has a light (ID 3).
If you have all five IDs of lightboards, the lowest one (ID 1) would always light-up even if the force sensor was not experiencing any load. Sometimes you want this, but for today's project, I wanted the bag to be unlit when it was on the floor, so I used IDs 2-5.
The other components of this projects are the force sensor itself, and the custom sensor board.
Everything on the schemer bus needs to have a microcontroller, because they all communicate in a custom protocol that delivers power and data on only two wires.
The other components of this projects are the force sensor itself, and the custom sensor board.
Although you can get boards with light, temperature and touch sensors built in, for this project you need the custom sensor which allows you to connect other things - like this force sensor.
Step 3: Arrangement and Sewing Data Lines
The force sensor should be attached at the shoulder where it would experience the most pressure. Then the lights cascade down, with the battery holder at the end of the line.
The wiring is really simple. Unlike projects with other type of wearable boards, you only need two lines of conductive thread: data and ground, which go to all the lightboards and sensor.
Basically, hook all the dots of schemer, lightboards and sensor together, in one run. This connects all their data pins.
The last picture illustrates how you just need one piece of thread to connect all components together. That's what cool about the schemer bus: simplicity.
The last picture illustrates how you just need one piece of thread to connect all components together. That's what cool about the schemer bus: simplicity.
Step 4: Minus Connections for Lightboards & Sensor
Normally, I'd start sewing the minus connections now, but to make things easier (in this case), I sew down one of the inputs to the custom sensor. Leave a long piece of thread, since you'll connect it later to the force sensor.
Next, connect all the minus holes. On schemer, when you align it with the + sign up, the bottom 2 holes are all minus. For lightboard, the minus hole is opposite the dot hole, and for sensor, it is the bottom hole when the text is right-side-up.
Step 5: Battery Power
Now sew down schemer's + hole to the right side of the battery holder, then sew down schemer's - hole to the left side of the battery holder. I really like how easy this is. I'm not just saying this, but for wearables, wiring should always be this easy.
Two reasons I'm working on the schemer system:
The first:
It hit me some time ago that the traditional ways of hooking up electronics are a big roadblock to broader acceptance of etextiles: for educators, for mass manufacturers, for DIYers, for artists, and so on.
On a circuit board, you can route all the wires carefully so they don't touch. This is easier to do on a hard, fixed, 2D circuit board where you have the luxury of routing wires under and above the circuit board.
However, in the case of sewing, the thread is stitched to both sides of fabric, making this kind of routing more difficult; it's in 3D; and the whole fabric moves; so it becomes a mess really quickly. We need to rethink how to electronics and fabric may better coexist, and I think simplifying and reducing the number of connections is a good way to start.
The second:
Programmable artifacts still require a relatively beefy computer (I'm talking about the need for hardware ports, software, keyboard, etc. and not processor specs). If I have a wearable thing and I need to change its behavior, the typical way would be to sit down, fire up an editor or development environment, type, compile, find a USB cable, download.
Wearables give us a tremendous opportunity to rethink this very premise. Why can't we program directly from our phones for instance? This would be awesome especially for quick changes. Could we download a program embedded in a painting? If I was a museum curator, could I make my visitors' light-up shirts behave differently depending on the exhibit? What would that mean? How would that be possible?
I feel that web-programmable wearable buttons are a good - though tiny - step in enabling a more pedestrian approach to etextiles.
Two reasons I'm working on the schemer system:
The first:
It hit me some time ago that the traditional ways of hooking up electronics are a big roadblock to broader acceptance of etextiles: for educators, for mass manufacturers, for DIYers, for artists, and so on.
On a circuit board, you can route all the wires carefully so they don't touch. This is easier to do on a hard, fixed, 2D circuit board where you have the luxury of routing wires under and above the circuit board.
However, in the case of sewing, the thread is stitched to both sides of fabric, making this kind of routing more difficult; it's in 3D; and the whole fabric moves; so it becomes a mess really quickly. We need to rethink how to electronics and fabric may better coexist, and I think simplifying and reducing the number of connections is a good way to start.
The second:
Programmable artifacts still require a relatively beefy computer (I'm talking about the need for hardware ports, software, keyboard, etc. and not processor specs). If I have a wearable thing and I need to change its behavior, the typical way would be to sit down, fire up an editor or development environment, type, compile, find a USB cable, download.
Wearables give us a tremendous opportunity to rethink this very premise. Why can't we program directly from our phones for instance? This would be awesome especially for quick changes. Could we download a program embedded in a painting? If I was a museum curator, could I make my visitors' light-up shirts behave differently depending on the exhibit? What would that mean? How would that be possible?
I feel that web-programmable wearable buttons are a good - though tiny - step in enabling a more pedestrian approach to etextiles.
Step 6: Testing
Insert the battery, plus side facing you. You should see schemer doing its initial sequencer pattern, unless you've reprogrammed it.
Step 7: Force Sensor
Now it's time to hookup the force sensor. This type has less resistance as more pressure is applied to it.
Choose a place where it's likely to experience the most pressure when you carry your bag. The heavier the bag, the more pressure in this spot.
Connect the two pins on the force sensor to the two side holes of the custom sensor board. It doesn't matter which one goes where.
Once sewn in, you can attach the sensive bit to the soulder pad. Reprogram using the schemer interface to get your desired response. See the video at the top of the page, or go to the introductory page to learn more.
