Someone in Apeldoorn, the Netherlands has set up a small network of five Bluetooth sensors. The sensors are placed in different locations in the city, as said on the projects homepage “with friends and family”, to keep it a low budget project. Each location has a simple USB Bluetooth adaptor connected to the internet. All the information is then stored in one location in a database.
The sensor will pick up mobile devices with Bluetooth turned on. They are identified by a unique MAC address. Through the network of sensors within Apeldoorn it is then possible to roughly track individual devices.
The amount of data collected from just five locations is quite a lot. Within the first four weeks the network registered 15’000 unique devices.
As it occurs, some devices are picked up by two or more sensors and are therefore reveal information about movement within Apeldoorn of individual devices. Data from one sensor over a period of time reveals a picture of the usage pattern of the area it covers. An example from the project homepage at bluetoothtracking.org.
Below you see the statistics of the Apeldoorn Driehuizen Bluetooth scanner. The location of this scanner is near a couple of office buildings. You can clearly see the early morning and late afternoon traffic it even shows that they usually go for a walk after lunch. One afternoon peaks at 12, Friday afternoon, and this is because many people take Friday afternoon off.
Image taken from Bluetooth tracking - Weekday activity
Dutch people like to enjoy long weekends and often take Friday afternoon off.”
This simple chart visualizes how working hours create a pattern in everyday movement. The chart only represents on e week, but every week is most likely the same as the pattern is repetitive.
The University of Bath has, it was revealed by a Guardian article on Monday July 21 2008, undertakes a very similar study. For this study the university has already three years ago, installed 10 Bluetooth scanners to capture signals from mobile devices. The data is used to study how people move around cities. The project is called Cityware and, similar to the previously mentioned example based in the Netherlands, stores the data centrally to allow analysis.
On the Cityware project home page the team publishes a map of data collected back in 2007. During a day on 9 different locations, 6 time sessions have been scanned.
Image taken from: Cityware