Virtual Traffic Control

Traffic control is used to control the cooperative behaviour of road users. At intersections, this is usually done by means of light signalling systems (traffic lights) in order to prioritise the right of way for road users in an appropriate manner and to balance interests in the different directions.

With connected, automated driving, the possibilities of control go far beyond the simple, blanket switching of traffic lights. Through connectivity and the automated exchange of information, vehicles can be controlled individually and in a more diverse way for improved, cooperative maneuver planning. However, this requires suitable algorithms and software to determine the best individual cooperation behavior in the respective situations. This results in a virtualization of traffic control, which takes place in distributed ECUs and computers. The traffic control system is therefore no longer necessarily linked to a traffic light and can also be applied to previously unregulated traffic nodes and sections of the route.

Challenges

• The additional possibilities of vehicle control and maneuver planning greatly increase the complexity if one really wants to leverage the advantages that can be achieved by vehicle automation and connectivity.
• Distributed algorithms do not necessarily lead to emergent improvements in overall traffic. On the contrary, cooperation strategies must be specifically designed and validated using sophisticated methods.
• Since these are safety-critical systems, the algorithms and software functions must be defined and safeguarded explicitly for all possible and conceivable states, interaction patterns and maneuvers. This results in a comprehensive and complete redefinition of traffic rules and regulations.

Solution Strategy

The solution for developing and safeguarding new control strategies for optimized virtual traffic control consists in a coordinated application of the Integral Development Process of ANDATA for all of the VERONET vehicles, node, line and net control modules. This also automates the development of control strategies for any road topology, traffic situations and road users, so that comprehensive and complete coverage can be achieved with self-learning and self-organized algorithms.