System Dependability
Topic outline
-
Defined as "the ability of a system to perform when and as required" (IEC), Dependability is also the science of ensuring this ability and acquiring justified confidence in it. Linked to risk management and often summarised under the term RAMS (Reliability, Availability, Maintainability, Safety), it has become essential in industrial fields (energy, transport, etc.) where failures can have major consequences for people, the environment, the economy and, more broadly, society. To provide scientific arguments for the design of new systems, we invite you to enter the world of operational safety: "know what is" (feedback) to "predict what will be" (methods and tools for analysis and calculation developed in this module) in order to "improve it as much as possible" (decision support to choose the best compromises).
-
Unité 0. Initiez-vous à la sûreté de fonctionnement
Cette Unité fournit une initiation à la sûreté de fonctionnement.
Unité 1. Contexte, concepts et démarche générale
Cette Unité fournit une présentation générale de la sûreté de fonctionnement, son contexte et son histoire. Elle introduit ses concepts de base (variables aléatoires, notions et paramètres probabilistes fondamentaux) puis la démarche générale avec ses étapes et les outils associés ainsi que les techniques simples de base (APR, HAZOP, AMDEC, Arbre d'événement).
Unité 2. Introduction aux analyses quantitatives, données de fiabilité et approche Markovienne
Cette Unité introduit les méthodes quantitatives (aperçu et typologie). Elle montre ensuite l’importance du retour d'expérience dans l'estimation des données de fiabilité. Elle expose les processus stochastiques puis l'approche markovienne très efficace pour illustrer les différents concepts et évaluer la sûreté de fonctionnement de petits systèmes complexes dynamiques.
Unité 3. Approche booléenne
Cette Unité applique les concepts Booléens aux diagrammes de fiabilité et aux arbres de défaillance : analyse qualitative (coupes minimales) et calculs probabilistes (diagrammes binaires de décision). Elle introduit ensuite la variable temps, étudie la combinaison avec les processus de Markov et compare les calculs de disponibilité (faciles) aux calculs de fiabilité (difficiles).
Unité 4. Simulation de Monte Carlo et Réseaux de Petri
Si les techniques abordées dans les trois Unités précédentes sont bien adaptées aux systèmes formés de composants indépendants, leur mise en œuvre ne supporte pas la complexité comportementale. Pour y suppléer, deux techniques complémentaires sont proposées dans cette Unité : la simulation de Monte Carlo associée aux réseaux de Petri stochastiques.
-
The first unit, "Initiez-vous à la sureté de fonctionnement", was created by three teacher-researchers from the Institut National des Sciences Appliquées Centre Val de Loire.
Frédéric Kratz is a Professor specializing in reliability analysis and predictive maintenance. He is affiliated with the PRISME research laboratory, where he leads the Engineering of Dependable Systems team within the Control Engineering division.
Vincent Idasiak is an Associate Professor specializing in reliability and safety analysis within Model-Based System Engineering (MBSE). He is affiliated with the PRISME research laboratory, focusing on enhancing the reliability of embedded systems through robust design methods. He promotes bridging MBSE and MBSA models. Since 1997, he has contributed to the creation of the Industrial Risk Management (MRI) and Computer Science and Technology (STI) departments at ENSIB, now INSA-CVL. He has been leading the Industrial System Risk (RSI) training for 25 years
Yunhui Hou. After a PhD at the Université Technologique de Compiègne (France) and a post-doc at the Commissariat à l'Energie Atomique (France), Yunhui Hou has been appointed Assistant Professor at the INSA Centre Val de Loire (France). Her teaching and research interests relate to systems dependability. She is responsible for the "Environmental Risks" training in the "Industrial Risk Management" department.
Jean-Pierre Signoret holds a master's degree in physics and has been actively retired since 2009. He began his career at the CEA (Commissariat à l'Energie Atomique) in nuclear safety, before continuing at ELF and TOTAL in the field of operational safety of safety and production systems. Responsible for both R&D and operational studies, he has been involved in the development of methods and tools for analysing and calculating operational safety for over 50 years now, and in the development of the GRIF software package, which has been implementing this expertise for over 40 years. A former Chairman of ESRA (European Safety and Reliability Association) and Vice-Chairman of ISdF (Institut de Sûreté de Fonctionnement), now IMdR (Institut de Maîtrise des Risques), he is the author or co-author of numerous publications and books, including "Reliability Assessment of Safety and Production Systems" (Springer), the basis of the first 4 Units of this module. Active for more than 20 years in standardisation, he chaired the AFNOR/UF56 committee (mirror of the IEC/TC56 committee on Dependability) and was project leader and member of several international ISO and IEC standardisation committees (functional safety, Dependability techniques, data collection, production availability, etc.). A member of TPA (TotalEnergies Professeurs Associés), he lectures at various universities around the world.