Reusing FIDES knowledge in the MéDISIS Dysfunctional Behavior Database

Robin Cressent (1) , Vincent Idasiak (2) , Frédéric Kratz (2)

1 PRISME - Laboratoire Pluridisciplinaire de Recherche en Ingénierie des Systèmes, Mécanique et Energétique [2008-2013]
2 IRAUS - Département Images, Robotique, Automatique et Signal [Orléans]

For 3 years our research team has applied on different industrial projects our MéDISIS methodology. The MéDISIS methodology aims to ease the integration of dependability analysis through the system engineering process. MéDISIS represents now a complete framework described through various publications. Currently, the use of MéDISIS relies on a functional system model in SysML, the Dysfunctional Behavior Database and processes of translation from SysML to FMEA, Altarica DataFlow, AADL and Simulink to perform dependability studies. Until recently, the DBD was a tool dedicated to MéDISIS. Consequently, the DBD was only designed to be updated through dependability analysis of project implementing the MéDISIS methodology. We describe in this article how we can complete the DBD with knowledge from other sources such as reliability repository. The main source we have studied is the FIDES guide. We highlight how the meta-model of our DBD and the model of the dependability database allows to connect them. We describe the process leading to failure law definition and failure rate calculation in the FIDES guide. It matches perfectly several concepts that are associated to the description of failure modes in the DBD meta-model.

Mots clés: Model-Based System Engineering; SysML; FIDES guide; Reliability database

Domaines: Automatique / Robotique

Robin Cressent, Vincent Idasiak, Frédéric Kratz. Reusing FIDES knowledge in the MéDISIS Dysfunctional Behavior Database. PSAM 11 & ESREL 2012, Jun 2012, Finland. pp.02-FR3-3. ⟨hal-00745670⟩