Intelligent Transport Systems are currently being developed in many different industry sectors. These developments range from highly automated land vehicles, robots for mail delivery, agricultural drones up to ships automating vehicle ferry operations or automating the transportation of oil from the corresponding platforms.

Virtual drivers are a big challenge for implementation of these systems, and there is currently much activity in this area. But this is not the major challenge;​ which is making those systems safe and reliable. The following article shows an approach to realize safety and reliability of Intelligent Transport Systems by separating the functional components into a driver model with limited safety and reliability, and an additional safety layer. In this approach, the driver model takes care of putting the required application case into practice and tries, similarly to a human driver, to continuously optimize the driving task. It is also possible to use training programs in productive operations for such driver models.

The driver model is supported by a static safety layer. This safety layer implements all safety targets that have been defined in the development phase and ensures that all safety targets are continuously being adhered to during the operation. This article shows an overview of the relevant safety targets for Intelligent Transport Systems and demonstrates strategies for implementing the security layer.