## Gesture Steered Turing Machines

A new astlab project, which comes closer to realize something which I have carried around in my head for now almost ten years.

A Turing machine is an abstract formulation of the underlying modes of operation of any classical computer. That is in principle you can remodel any action on your computer as a Turing machine. The Turing machine in the video is a 2 symbol Turing machine, which means that the program prints only black and white and not more colours. In the video you can see a diagram with 6 “nodes” or “dots”, which encode 6 states. So the video shows a “six state Turing machine.” Moreover the program starts with an all black register or as some people say “tape” (bottom right corner). Starting with only one color is also the initial register state in the Busy beaver game (history of busy beaver games). In the video the tape is however periodic, which means it is intrinsically glued together at the ends.
In the first step a single white box, or blotch or pixel or whatever you want to call it, is printed on tape. You see that newly printed tape underneath the original tape. The graphics with the six nodes describes how the program is performed, i.e. which program step is to be performed after which and which colour has to be printed on the tape and where. So the diagram is a visual representation of a Turing computer program. Since it can be altered via the hand gestures this means that one can (re)program the Turing machine via gestures.

The hand gestures are read in via a sensor-laden glove, which I had assembled by following Hannah Perner-Wilson’s instructions. The glove exists also in a professional version, offered by the company mimugloves. It is currently sold out, but you can preorder. Making a glove yourself following Hannah’s instructions is – depending on the cost of your work-time – likely way cheaper, but the old DIY version is likewise way less capable than the professional mimugloves. In particular unfortunately the bluetooth connection is not very stable (thats why I use cables in the video), which seems to be mostly due to the driver though.

The above video is more or less a proof of concept, so in particular it has been filmed in one (and only) take without script. I currently don’t have time for fancier presentations. The Gesture Steered Turing machine is thought to be the first project in a collection of projects called BODYCONTROL. More full body interaction and higher level programming languages are more or less in the making.

My personal motivation to currently push for more realizations of the concept of “programming by full body performance” is that the past year I had to do a lot of (partially rather monotonous or otherwise dull) tasks in front of a computer screen, which had not the best effects on my health. I just say eyes, weight, shoulder tension etc. Likewise I heard similar things from people doing computer screen work in the Berliner Verwaltung – despite the fact that employees there, as I learned, get some free sports classes etc. in order to mitigate the problem. But there may of course be other motivations to explore this approach of “body programming or body control” a bit more in-depth like for kinesthetic learners.

Here some examples of complex body gestures in dance:

- Finger tutting dubstep hand gestures by Pnut on youtube which would definitely be a challenge to any screening system – even to the mimugloves
- Rather precise arm gestures by Kai Sosceles in a video on facebook (via Maike Bartz) (for comparision a more freestyle popping/turfing performance: Turf vs Terry on youtube)
-a Bharata Natyam dance by Savitha Sastry on youtube using quite a couple of mudras in a dance which displays a Jathi in three different speeds (i.e. this is a tri kala Jathi).

The word Kal by the way means also to count and may – according to the Monier-Williams Sanskrit-English Dictionary, 1899 – be found in the latin word calculate (in fact “Kal Kula” is “count multitude”) and -as a nonexpert- I think Kal may eventually also be found in the german word Zahl.

comments in german, french and russian will be translated into english.
you can use LaTeX in your math comments, by using the $shortcode: [latex] E = m c^2$