Tracking Dynamics in Concurrent Digital Twins

The availability of machine-generated data for the management of complex systems enables run-time technologies for diagnosis, predictive maintenance, process control, etc. that find their apex in digital twins. Such model-based replica of cyber-physical assets represent system elements and their behavior within their environment, which is often dynamic. These dynamics of a system's environment can render the underlying model unfit w.r.t. the changing reality and thus cripple the whole approach. We provide the means to detect such a transgression of the operational space of digital twins and similar technologies using a novel combination of probability-of-findings calculations with established process control methods and localize necessary updates to ensure efficient model maintenance.