Künstliche Intelligenz zur Verbesserung der Sicherheit von Tunneln und Tunnelleitzentralen
As a result of increasing traffic volumes and based on the European Directive 2004/54/EG on minimum safety requirements for tunnels in the trans-European road network, the use and complexity of monitoring and control tasks as well as the required information and communication technologies (ICT) in tunnels increase continuously. In order to cope with this complexity, in recent years tunnel monitoring and control duties are centralized in tunnel control centers. Due to the increase in ICT, the number of digital points of contact is growing. Thus, it is getting even more challenging to ensure operational and, as a result, personal safety. At the same time developments in the area of Mobility 4.0, particularly Car-to-Infrastructure communication (C2I), will provide additional data in real time, such as information about the current traffic situation or vehicle conditions. Taking into account innovative safety mechanisms, this data can be used to increase traffic safety as well as the availability of road tunnels.
On the other hand networked mobility implicates upcoming risks and efforts in the evaluation and use of new information. Hence there will be a further increase in the operator’s workload in tunnel control centers. In case of incidents or serious events operators have to make decisions immediately and therefore are faced with high stress. Using the available data with regard to C2I-communication in the medium term, concepts have to be developed to use them in a target-oriented manner. This is where the implementation of weak AI offers a promising approach. It supports operators in assessing safety issues, decision-making in case of incidents and predicting dangerous situations. The aim is to develop concepts which allow the training and systematic optimization of operator actions in case of incidents.
In a network of tunnel operators, tunnel suppliers, rescue services, experts in the field of risk assessment and scientists from the area of IT-security and AI a concept is to be developed early on to cope with the upcoming challenges and to improve the safety-related situation assessment of tunnels and tunnel control centers. AI shall make use of data based on already existing tunnel sensor systems as well as on future C2I-communication. Furthermore, possibilities of influencing traffic based on the Infrastructure-to-Car communication (I2C) are reviewed and general conditions for realization are established.
- Partner: AIT Austrian Institute of Technology, ILF Consulting Engineers Austria GmbH (Koordinator), Universität Wien - Rechtswissenschaftliche Fakultät, Österreichischer Bundesfeuerwehrverband, ASFINAG Maut Service GmbH
- Project duration: 01/2021-12/2023
- Funding: KIRAS Sicherheitsforschung - Ausschreibung 2019/20, KI, Länderübergreifend