Testing a digital level crossing warning system for road traffic users with a naturalistic driving study

Kurzfassung

The points where roads intersect with railroads, called level-crossings, have traditionally been considered dangerous due to the excessive severity of accidents that have occurred involving all kinds of road vehicles and trains. They are most frequently protected with regular safety systems like barriers of light signals, however numerous level-crossings still remain unprotected. In addition, the maintenance status of level crossings differs strongly across Europe. As a consequence, some level crossings pose serious safety threats to road- and rail users, if the level crossing safety infrastructure does not exist or is not properly maintained and operated. Especially at level crossings in urban areas of municipalities in developing countries in Europe, it is often overly costly or even impossible to protect a high number of level crossings and maintain them on a regular basis. Digitalization, the huge coverage of mobile telecommunications networks and the penetration of smart phones and mobile devices in society offer a starting point for alternative, hi tech safety measures, that promise to be cheap as well as effective. Such a system has been developed and tested in Thessaloniki, where a mobile device elicits an auditory as well as a visual warning whenever the user drives in close proximity and heading to any of the 30 level crossings located in the suburbs of the city. Some trains have also been equipped with Global Navigation Satellite System (GNSS) sensors which record and transmit highly accurate spatiotemporal data, so when they approach the level crossings the mobile users also receive the estimated time of arrival on their smart phones and tablets. Three taxis were equipped with a naturalistic driving platform, consisting of cameras and a GPS sensor, to track the behavior of the drivers whenever they approached level crossings. The equipment was installed taking into consideration both the safety of all passengers and the privacy of customers. Only the drivers and surroundings of the vehicles were recorded, for a prolonged period that lasted more than two months and resulted in more than 1TB of relevant video data for further analysis. During the first month the alert system was not activated, in order to collect baseline data. The safety relevant behavior of the drivers in the context of level crossings before the implementation of the assistance system was compared to their behavior when using the system.