Flitter is a stand-alone social monitoring device that networks with the Nokia N59 to log positions of the head and proximity of other speakers. It is designed to record head gestures within the context speaker proximity. Flitter is composed of an accelerometer, ultrasonic range finder, blueSmirf bluetooth unit, and an arduino microcontroller all housed in an elegant hair piece. It is a child of the cocktail dress (Prototype and tutorial.)
While the cocktail dress was not a complete bust, its last iteration made it clear that I was constructing this piece in the wrong manner. I love dresses; they are a single article of clothing that can cover the entire body and putting electronics in them can be so much fun. Yet, this is not the piece to put in a dress. The accelerometer is, by nature of what it is monitoring, already mounted on the head. Testing the proximity sensor showed that the lower it was on the body the more noise it would encounter from hands and laptops, as well as being less sensitive to conversations that one was not facing with their whole body.
Therefore, I have transferred the electronic components of the dress into a stand alone hair piece, which I have sarcastically dubbed “Flitter.” The new environment was designed specially for this Saturday’s Gala event, celebrating ITP’s 30 year anniversary. I intend to be logging live data at the event, and it seemed that any hardware should be suitably attired for such a major event.
The last version of Flitter (The Cocktail Dress), networked with the computer. Thus the results were logged to a text file through Processing. The problem with this setup was two fold, one sprouting from the other. While the laptop is quite handy and mobile as far as computers go, it is still too bulky to smoothly integrate with a wearable piece. Thus, it has to either be housed in a bag and carried around or left in one location. Leaving it in one location brings into play the second problem: range. While the bluetooth connection allows you more mobility than a USB connection, the range is still limited to a couple yards, which, once again, is rather impractical. The solution is to resort to either a network with a stronger signal or a smaller, mobile logger.
I went with the latter, borrowing a Nokia N59 to receive bluetooth serial information and log the information on the SD card. The logging software is written by Robert Carlson and can be found on his github account.
Along with the new recording device, I have also updated the power source, using two 3V coin cell batteries connected to the arduino with a switch to turn the device off an on. This now frees it from the 4 AA batteries and container that were earlier required and enables turning on and off in a much simpler manner.
I will hopefully have some fun pictures with Twitter after the gala is over.