« Guide to salutations | Main | Reflections on July 4th »

September 20, 2004

Experiments with Crowds

Lately I've been trying to come up with a simple way of inferring crowd size using simple electronic sensors. While sitting around in a lab meeting, we came up with the brilliant idea that since some people have Bluetooth-enabled cell-phones or PDAs, then one could come to the conclusion that the more Bluetooth devices that you detect, the more people there are. In fact, in England, "toothing" is a popular (if unwholesome) means of finding random people to talk with, because in a crowd of people on a train, one can usually depend on finding some with a bluetooth-enabled device to communicate with.

So, lugging my trusty laptop and a serial-to-bluetooth device on loan from my lab, I decided to go to every single Starbucks that I saw on my way through Manhattan. As you might imagine, in a city as dense as New York, there are quite a few branches of Starbucks. This gave me a dependable pool of establishments with a similar layout and a clientele representing a consistent demographic.

What did I learn? First, I learned that there aren't any branches of Starbucks on 6th Avenue between Greenwich Village and Chelsea. The other thing I learned was that hardly anyone has a Bluetooth-enabled cell-phone, these days. Seriously. I detected only 2 Bluetooth-devices the whole day, both of which I was only able to detect once, briefly-- I assume because they belonged to people walking by outside the Starbucks, rather than inside the coffeeshop. So, back to the drawing board when it comes to inferring crowd size.

I also learned that while Starbucks' coffee might be awful, their tea is pretty good.

More detailed results (yes, I was able to make some rather subtle conclusions) will follow, soon.

Posted by Dean at September 20, 2004 10:43 AM

Trackback Pings

TrackBack URL for this entry:
http://www.christakos.com/cgi-bin/mt-tb.cgi/33