A Digital Twin Method for Automated Behavior Analysis of Large-Scale Distributed IoT Systems

Abstract

The behavior of a large-scale distributed IoT system is often hard to verify and validate. The reasons include: 1) the specification is often unclear, ambiguous and incomplete resulting in misunderstandings and undesired behavior. 2) It is almost impossible for a human to reason about the correctness of a system consisting of thousands of components. 3) It is very hard to observe all related components when trying to solve a problem because the system is geographically distributed over large areas. A digital twin capturing the system operational behavior will be of great help to assist a human in detecting behavioral anomalies and reasoning about root-causes. This paper proposes a method to develop digital twins for automated behavior analysis of large-scale distributed IoT systems. We present a real-life use-case of a smart office lighting system for which the method was successfully applied. The developed digital twin was used for anomaly detection and reasoning in a semi-automated root-cause analysis (RCA) approach.