Brussels / 2 & 3 February 2013


Vehicular traffic estimation through bluetooth

We present a traffic monitoring prototype powered by a raspberry-pi that leverages on techniques such as the detection of signals as Bluetooth. Namely, signals spread by vehicles passing on the roadway are revealed by the battery powered boxes installed on the roadside.

The entire system is proudly powered by only free and open source softwares, at the probe side upon a healthy OS level orchestrated by the Debian OS runs the bluetooth inquiring tool while at the server side the traces are gathered and analyzed by a python back-end developed with the web2py web framework.

Given the growing and serious issue as the traffic jam and the continued reduction of the budget that municipalities have to deal with a sane and open source system for monitoring the traffic trends could be a starting point not only for cutting the expenditure down but also to develop an homogeneous monitoring infrastructure.

The aim of developing our prototype based on signals as bluetooth stems from an interest of proposing an open source solution in the field of monitoring vehicular traffic. Moreover the idea is that the same solution can be applied by different municipalities without any significant effort in customizing and adapting the source code for their particular requirements. Nowadays the bluetooth technology is available on several vehicular fleets, it is a matter of fact, it is widely used as the main in-car short range point-to-point communication standard for info-entertainment and phone headset. Its adoption in new vehicles is growing, as a result we expect that the number of cars monitored will increase in the following years. This positive growth can only strengthen the use of this technology. In particular, based on empirical tests carried out in the city of Bolzano, the number of cars detected are at least the 25% of the total traffic flow, with an average of 30%, and peak of 43%. This figures are proof that it is possible and worthwhile to predict the traffic trends with this approach. In addition, by detecting bluetooth signals the system is not only able to project an estimation of the total number of cars passing through the monitored area but also to compute the travel time that a vehicle took to pass through two different places monitored by the probes. Namely, the former is an estimation chart generated as a function of the percentage of cars equipped with bluetooth while the later can provide input data for the so called in the Intelligent Transportation System (ITS) domain origin/destination matrix, which is used by traffic engineers to feed complex traffic simulation models that compute the traffic flows distribution over the entire road-network.


Paolo Valleri