[view brochure in Dutch]

[view paper (Chan et al, TRB 2008)]

[Nieuwsbrief Juni 2009]


To manage and monitor traffic the Dutch Department of Transport (DoT) relies on an extensive network of traffic sensors. Data produced by these sensors are processed by a diversity of applications that each perform a specific task, like automated incident detection, speed advice, queue length display or preparing radio broadcast information. Because these separate applications rely on each others output and function under the DoT’s responsibility we speak of DoT’s corporate traffic information chain.


The complexity of these systems and the fact that tasks are distributed over different applications and departments makes insuring the correct technical functioning of the information chain difficult, both from a technical ad a organisational viewpoint. Therefore DoT has decided to develop a system aimed specifically for this task: the DaVinci system.

The Da Vinci system is not aimed at replacing or integrating any of the existing systems, but merely to supervise them. To this end Da Vinci contains functions for both automated detection of errors and functions for visualisation of traffic data. The automated detection of errors presently relies on three models: a logical model that evaluates binary expressions or combinations thereof, a time series model that scans data for outliers, discontinuities and abrupt structural changes, and a spatial model that checks the consistency of data from different sensors. The visualisation tools enable presenting sensor data such as flows and speed and sensor metadata such as sensor location and road layout. A key feature is the drill-down capability: the user can view corresponding data from different systems simultaneously and back-trace the information chain.

Da Vinci is an example of how validation of an entire information chain can be achieved. It will significantly reduce the costs of application maintenance and development and improve the overall quality of data, thus enabling new and valuable services based on these data.


Press "Next" to view some screenshots