Thursday, July 23, 2009

Algorithm

I'm in Siena, a medieval walled city. This place is exactly like the kind of mazes you find on children's placements at diners all over the US. (Maybe internationally too... I haven't ordered a kids meal here...) With this fact in mind it made sense to me to try and "solve" Siena using maze solving algorithms.

Maze Solving Algorithm

Now, I'm finding that when I say the word algorithm, people around me seem to just tune out. So let me clarify really quick. By algorithm I don't mean a large convoluted series of equations. I mean a very simple, specific, and strict set of rules. With Siena those rules are as follows:

1) Choose two gates, one to start at and one to finish at.
2) Choose a hand, left or right.
3) Keep that hand on a wall at all times. [Note: you don't actually have to walk touching a wall...that would be silly...but your path must be determined by the ability to do so. IE, if you choose your right hand, every time you can make a right turn, you must.]
4) Treat all gates that are not your final destination as though they are closed, with a wall bridging them.

Because you are starting and ending along city wall, you are guaranteed to get to your destination. It won't be the most direct route, but that's kind of the point.

Along the same lines, but with a less clear destination, is my second method - RPS.


RPS - Random Positioning System

RPS does not guarantee that you will get anywhere, and it is almost guaranteed that you will become lost within the city. After I get back to the states RPS will become an actual ipod touch / iphone app, available for download and applicable to any city. For the time being though I'm accomplishing the same effect with a regular ipod and the following rules.

1) Load an ipod with 3 images, a left arrow, a right arrow, and a forward arrow. Put them in a playlist. Enable shuffle.
2) As you walk, at every interesection, restart the playlist. This will select a new image at random.
3) Follow the arrows.
[NOTE: This WILL get you lost. Pay attention to the route it takes you on so you can back track if necessary.]

Historic Background
Like most things, the basic ideas behind these methods of navigation are not entirely new. And it seems fitting that in the process of prepping for this piece I stumbled across Psychogeography (by stumbled I mean I wound up in its territory and was alerted of this fact by my professor Colin Ives)

In 1955 by Guy Debord defined Psychogeography as "the study of the precise laws and specific effects of the geographical environment, consciously organized or not, on the emotions and behavior of individuals." Guy Debord and the Situationalists used a map of London to explore the German countryside, using a restricted system to approach a nearlly unrestricted geography.

Using a map of Siena clearly wouldn't introduce any new experience to exploring the city, so instead I'm applying rules of mechanical exploration to a human system.

2 comments:

zacharyr said...

I love RPS. Perhaps it should have a slight bias towards getting you somewhere else? (http://en.wikipedia.org/wiki/Random_walk)

Andrew S. Parnell said...

Thanks Zach. I honestly prefer using the Maze Solving Algorithm, but it is more limited to Siena (and other walled cities.)

The current design of the RPS app does bias in that it will never direct you to go backwards, but it could conceivably put you in a never ending circle or right turns. The odds are slim though. (Maybe I'll throw a bias in... hrmm)

Thanks for the article on Random Walk.