| The increasing dependence of our societies towards Lifeline Networks (LN) increases their vulnerability. LN, which are entities that provide to the society the resources essential for their proper functioning (water, gas, electricity, telephone links, etc.), are highly dependent on each other. Thus, when a disturbance affects a LN, this may cause the failure of another LN whether because it uses its resource or because it is geographically located close to it. This can then cause the failure of other networks. These cascading failures are characteristic of a Domino Effect (DE) that could lead to total paralysis of the socio-economic environment.;The research presented in this paper focuses on developing and validating an Expert System (ES) which will eventually be integrated as the backbone of the EWS. The methodology used to develop this ES is composed of several distinct tasks. Initially, it is necessary to structure and simplify the Knowledge Base (KB) containing all the knowledge and expertise held by experts of the different LN and necessary to identify and anticipate DE. Then, an analysis of the needs expressed by the eventual users of the ES must undertake. The software MS Access was chosen as the platform for developing the ES. Then, the architecture of the tool was established and the ES was programmed. Finally, a phase of tests and validations of the ES was initiated to ensure that the needs of the future users of the system were satisfied.;Following the validation of the tool, the ES showed few limits. Some of them where due to the limited capacities of MS Access to generate complex graphics or functions. Moreover, it was necessary to develop new features to better meet the expectations of the future users, to increase data security and to better identify the interdependences between LN.;To conclude, the ES prototype obtained after the works presented in this paper meets the overall needs expressed by the users. Although a lot of work has still to be done in order to ensure a proper functioning of the system and a complete satisfaction or the future users, it represents a major advance because it constitutes the basic architecture to be used for the definitive ES. To best achieve the objectives of the project it is necessary to continue to work with the experts of LN to improve the ES.;The necessity of protecting LN and, thus, to reduce the vulnerability of our societies and the associated risks of failure are incentive to expand our knowledge of interdependencies between LN. In this context, the Centre risque & performance (CRP) of the Ecole Polytechnique de Montreal has led since a decade important researches in this specific area of study. From 2005 to 2008, the CRP was able to develop, along with multiple industrial and governmental partners, a simple, but yet very effective, methodology for identifying and anticipating DE between LN. Following this result, the CRP is currently conducting a research project aiming at developing an Early Warning System (EWS), allowing real-time management of interdependencies between RSV. |
