System Self Diagnosis for industrial devices

Downtimes of failed devices in an industrial plant must be kept to a minimum to achieve high system availability. Since failures are often caused by transient hardware- or software faults, a well-defined System Self Diagnosis (SSD) functionality is an important feature for the effective long-term operation of industrial plants. To achieve effective SSD techniques for network infrastructure devices, a root cause analysis of past failures was conducted and a fault model derived. A set of error detection methods was derived based on the checked root cause indicators. For productive deployment, an extensible SSD framework in form of a rule based system was designed and implemented to be used as an embedded software tool throughout the whole product lifecycle of an industrial device, achieving a highly efficient “Design for Testability” approach.