Friday, October 8, 2010

Typewriter to testubes - a concept shift

This week marks a big turning point in my project. I have realised that I will not be able to complete the typewriter concept within the time (and money) restrictions I have. Instead I have modified the concept to a more simplistic output, while still keeping the majority of the underlying concepts.

I will develop a system of using the CNC axis to fill test tubes, based on input from the web. In this way the tubes are more of a physical bar graph of activity. I toyed with many ideas involving logging live streaming data (eg. every time someone is killed in a particular manner, put a drop of blood in the corresponding test-tube. Morbid I know, but with lots of emotional impact which would strengthen my concepts) however I was unable to find any relevant data streams. It seems that because asynchronous content is fairly new technology, the companies that would output the data either haven't caught up with the available technology, or stream their data on a proprietary system that is closed to the public.

One idea that I stuck with for a few days was to have each test tube correspond to a country and provide a website that has unrelated content on it, but use the user's IP addresses to find what country they are from and put a drop in that corresponding tube. This concept relates to ideas about getting information off people without permission. I abandoned this idea because it was weak, and the link between the concept and the output was completely arbitrary.

Instead I have gone with an idea that is more closely related to my original plan, where users access a website and simply make the choice as to what tube they want to add their drop to. I will use the results of the experiment to attempt to make a comment on the thinking behind each user's choice. I will include a comment section on the website so that there can be some interaction between the users, which may be interesting to observe. This concept fits well with the test tubes aesthetic as it is in the spirit of an experiment.

I spent most of this week working on the software for the arduino, developing a system that makes the axis self calibrate by moving to one end until a limit switch is triggered, then to the other end til the opposing switch is hit, then returning the number of steps it took to the program. This allows me to consistently know the length axis, regardless of any physical changes I make to the system.

One of the biggest hurdles with the software was getting a system of Processing communicating with the Arduino microcontroller over the serial interface. The problem with the serial port is that it works both ways over the same line, so if two processes attempt to write to the serial port simultaneously the bytes merge as they are written, making neither program able to read the other program's message. To combat this technical difficulty I developed a system of handshaking where the processing sketch sends a command over, then listens for the confirmation of receipt message from the arduino before allowing any process to continue. The arduino also provides feedback on completion of an event, which is carefully written to only be passed when the processing sketch is not sending anything.

Tags: typewriter testubes arduino processing

←Newer: Testub es hardware and softwareOlder: CNC axis full motion working→