Predictive Runtime Simulation for Building Trust in Cooperative Autonomous Systems

Abstract

Future autonomous systems will also be cooperative systems. They will interact with each other, with traffic infrastructure, with cloud services and with other systems. In such an open ecosystem trust is of fundamental importance, because cooperation between systems is key for many innovation applications and services. Without an adequate notion of trust, as well as means to maintain and use it, the full potential of autonomous systems thus cannot be unlocked. In this paper, we discuss what constitutes trust in autonomous cooperative systems and sketch out a corresponding multifaceted notion of trust. We then go on to discuss a predictive runtime simulation approach as a building block for trust and elaborate on means to secure this approach.