Cost-Effective Hot-Standby Redundancy with Synchronization Using EtherCAT and Real-Time Ethernet Protocols

When designing and manufacturing industrial computer systems, many different issues must be considered. The most important of these are the temporal characteristics and real-time operation. Another factor is the distribution of the system, usually over a wide area, operation in harsh environments, and, finally, the system's dependability. When addressing the last issue, redundancy is commonly employed. This can be applied in many ways, e.g., by providing redundant communication links (buses and communication modules) and redundant sensors, input/output modules. This article considers systems that have computing unit redundancy, for which programmable logic controllers (PLCs) are often used. On the one hand, systems of this kind should be reliable and capable of real-time operations. On the other hand, they should be economically attractive. There are many solutions available on the market for the implementation of systems with redundancy. However, they are always rather expensive. Of course, the price is for a robust solution with dedicated hardware and many built-in features that are crucial in highly reliable systems, such as autodiagnostics. Nevertheless, such powerful products are not always required. This article presents an example of a cost-effective implementation of an industrial controller with hot-standby redundancy using regular devices and an Ethernet communication interface for the needs of synchronization of the computing units. For this purpose, EtherCAT and real-time Ethernet protocols were used. Note to Practitioners - This article is motivated by the problem of high cost of application of systems with computing units redundancy based on dedicated components, such as the latest S7-1500R/H Redundant System by Siemens or PACSystems Profinet System Redundancy by Emerson (former GE IP). The above solutions are powerful and sophisticated. However, sometimes, simpler solutions would suffice. This article shows how a system with computing unit redundancy could be built in a cost-effective way and, at the same time, provide all the basic features expected from systems with redundancy, i.e., fault detection, automatic switchover, and synchronization. All this is done with a common embedded programmable device and an open Ethernet-based communication network.