The Breathing Bag

Breathing Bag by Fiona Banner

Artist Fiona Banner asked me to assist her studio in making an art piece. I’ll describe the electronic process of making this build in the post, coding part was relatively straightforward.

Fiona’s assistant Chris Beauregard had made a rig which blew air in or out of plastic bag depending on how a switch was set.

The idea of the art piece was to make the Fiona’s classic catalogue bag seem to breath like lungs. The problem they were having was that even though the setup with fan mounted to the back of hole in the wall with fans facing oppisites sides worked for them, they couldn’t automate it without an electronic circuit. This is where I get involved and armed with and Arduino and a bunch of research I set off to make this bag breathe, so it could be permanently displayed.

The setup started really simply with one fan, an arduino, a transistor, a diode and two power supplies (one for the arduino and a higher 12 V to run the fans). This soon came to include both fans and a lot of cardboard and tape.  Silent PC fans were used to reduce the amount of noise the piece produced. One of the other things I was interested in is seeing if I could vary the power coming in and out of the fan using the PWM out pins from the arduino. this was creating an issue though, as the hard edges of the PWM especially at lower speeds were causing the fan to just turn on and off, I decided to put in some capacitors to see if I could smooth the signal coming in with these way too big caps (this turned out to bite me later).

After I was satisfied with my wire setup I bought an empty arduino shield so I could easily make a tight package. Since this needed to be on permanent display with no upkeep robustness was important. This also meant it needed to be mounted in some sort of box and we needed solid cabling. Chris made me a box with the mounts for the cable ports. I wanted the box to be hooked up to the power supplies and the fans in a way that was portable and extendible. I picked some popular cable types that could easily be bought in different lengths and replaced each one unique to prevent hooking things up incorrectly, I picked 3.5 inch jack, phono, 1/4 inch jack and DIN 5 (commonly used for MIDI). Note that DIN 5 was hooked up to an ultrasonic sensor which ultimatley scrapped from the project.

So this was it, or at least so I thought, one of the capicitors blew on the project and I had to go off to the gallery to fix it, so I quickly made my version two of this control box. One of the things I wanted to achieve is a more simple robust design that had fewer components so fewer things that could go wrong, so I went on to remove the capicitors. Removing the caps meant I wouldn’t be able to properly use PWM to gently ramp up and down the fans, but I found in practice that the interval that breathing worked over was so short it did not really make a difference. Instead I started to emperically time how and when the fan reached full power and when I would have to pre-emptively turn off the fan to get the desired reduced blowing speed to match up with natural breathing look for the bag.

I’m happy to say that my onsite surgery was a succes and the bag was up and running again.

After this Fiona asked me to make a few more so I also ended up making a version three, which I housed myself in a more efficient PSU box.

All in all I made six of these units, I had a lot fun making them.

Art piece.


Old Ascii art

Made with a processing sketch to draw ascii things in pleasing patterns.

image made with terminal drop

The birth of a realtime particle system

My old University project: Theia’s Aura gave me my first taste of programming with OpenFrameworks and putting together a realtime particle system.

First attempts at drawing with mouse and color.

Added opacity overtime and velocity, spread and birthrate controlled by mouse velocity.

Added color over life as well as randomised color fade and a cheat blur effect.

Heartbeat necklace

necklace testFor a long while now I have been keen on doing a wearable project. The intersection of art and the body has always interested me and since wearables are an increasingly popular categories companies like Adafruit have been making amazing strides in making these projects in a non industrial, experimental way. The problem is A: I do not have any fashion sense and more importantly B: I can not make Clothes. Enter Sandra Kaas, an old friend of mine and artistic allrounder: theatre, video, performace and most importantly fashion!


I started drawing up some plans, the brief was it needed to be managable and interactive in a way that would be easily understood. My first idea was to make a compass collar. I thought a fabric collar could be produced in a short time and it could be covered in lights all around with one color pointing north always and turning as you turned.

I wanted to use the flora, an alligator clip testArduino variant, as it seemed a great fit something wearable as well as these amazing rgb leds with their own drivers (which would make assembly easier), looking around the site I found this Accelerometer/Compass Sensor and I thought this could create an interesting interaction, I liked the idea of colour pattern that rotated itself north. Before going and meeting Sandra to put the things together I did some prototyping.

carboard prototypeThe module worked as expected but it was quite hard to test facing and colour effect (I was cycling through an RGB colour pattern depending on which direction I was facing with the RGB led on the flora chip itself), so I set out to make my beautiful cardboard collar prototype. I taped the chip to it.
cardboard prototype with chip
This still was quite unwieldy although I got some useful results from it finding

test 2

out approximate facing direction and how to translate a 3d vector for magnetic north in to a 2d one. I decided to try out the stainless steel thread which I was going to sew the sensor the flora and the lights together.

After this it was enough of my rough prototypes and I went to visit Sandra Kaas.

test 6

Test 7We discussed and she made some changes to the design taking beyond my functional prototype to give it actual shape, made it into more of a necklace than a collar. Putting together a nice looking fabric necklace, she started sewing on the flora  and the lights.

We made the colour pattern and tested the turning, I still had some tweaking to do. There were also some important lessons here, the thread had to very firmly be sewn on the copper copper connectors and the sewn electrical tracks needed to be very straight and firm as to not to cross connect wires when twisting and turning the fabric.

Test 9

Later as I was tweaking the code to make the colours change and turn to magnetic north Sandra came up with an idea connected to the theme of her own recent performances, couldn’t we make the necklace beat on the rythm of your heart? This was an interesting suggestion, I had never actively looked but I though their might be a heart sensor I could connect without too much trouble to an Arduino like device. As it turns out there is, a company called pulse sensor does exactly that. As I looked through the documentation it looked like it would be pretty doable in a short space of time, so I bought one and integrated it.

heart necklace earpiece

After some modifications to the old design and attaching the earpiece which through light and reflection measure bloodflow, I got together a first physically complete but still a prototype in terms of programming the lights.

I iterated over the programming and gave the lights some animation and more livelyness, I prototyped this in the awesome processing language for much quicker iteration. I finished up for now at least as I consider modifications to this design and new wearable ideas. Many thanks to my collaborator Sandra and all the awesome creative hardware and software to make it possible.