A Method to Evaluate the Dependability of DIBAS

Abstract This paper deals with a method to evaluate the a priori dependability of a Distributed Intelligence Base Automation System. Its aim is to give an estimation of the availability during the design step of a DIBAS. This method includes a descending step which allows to split-up the automation system into a system and subsystems which are associated with a function and sub-functions. Then an ascending step, which uses this previous functional modelling, allows to reconstruct the “working modes” of each subsystem. This last step is based on the working mode of material resources or in other words, on the knowledge of the functional and dysfunctional behavioural description. Concurrently, working modes truth tables define how the subsystems work together. A computing tool was designed. It facilitates drawing a DIBAS model and allows to do a consistency test, in order to estimate the availability of each subsystem and finally of the global system.