Optimized incremental state replication for automation controllers

Abstract

This paper presents optimization techniques for implementing software-based redundancy in industrial control devices. Initially, a brief survey of software-based state replication techniques with a special focus on their applicability in industrial control devices is conducted. The scalability, predictability and low latency of the technique are of particular interest in this case. Based on this survey, an analytic evaluation of different implementation alternatives is performed. As part of this analysis, a novel state replication algorithm is introduced. The approach uses support from the compiler or runtime environment to detect changes in the application state with very low overhead. This information is used to replicate the state of an automation controller in a redundant setup. Lastly, experimental results using a prototype implementation of the presented technique demonstrate that the proposed novel approach is able to perform state replication with constant overhead.