Integrating time-triggered and event-triggered traffic in a hard real-time system

In industrial communications, there is a high demand for flexibility as well as dependable and timely communication. There are two main paradigms for setting up a distributed real-time system, the time-triggered and the event-triggered approach. While the time-triggered approach is more rigorous and verifiable, the event-triggered approach is more flexible but harder to prove. In this paper, we present an approach of a distributed real-time system that follows the time-triggered paradigm but is able to operate in an event-triggered mode when the timing of the particular tasks does not stay within the initial assumptions. The system features an innovative on-the-fly monitoring mechanism based on slot violations. This way, the system stays operable while being able to inform about the mode change and therefore a possible problem.