Multi-Agents Model Oriented Safety in Maintenance (MAM-SM)

Hafida Bouloiz (1) , Emmanuel Garbolino (2)

1 ENSA - Ecole Nationale des Sciences Appliquées [Agadir]
2 CRC - Centre de recherche sur les Risques et les Crises

This paper proposes an agent-based simulation framework for the development of a decision support system for occupational risks management in a maintenance task. The proposed model is defined as a Multi-Agent system oriented Safety in Maintenance (MAM-SM). This model aggregates many agents, where architecture includes agents Supervisor, Resource, Machine, Environment, Reasoning, Task, Control and Agent Capitalization. Based on a multi-agent simulator, the objective of the proposed approach is to account for the complexity of the maintenance task for better analysis and understanding of risks. It allows orienting the actors to the best decisions in order to minimize risks that may arise. The method is applied to two case studies. The results show that this model can express the behavior of each agent and also the performance of the whole system. In particular, the results demonstrate that the maintenance tasks can be controlled to avoid an accident.

Mots clés: Multi-agents system; safety; complexity; risks management; decision support

Domaines: Gestion des risques et des crises

Hafida Bouloiz, Emmanuel Garbolino. Multi-Agents Model Oriented Safety in Maintenance (MAM-SM). International Journal of Computer Aided and Engineering, 2014, 24 (2), pp.1117-1129. ⟨hal-01269940⟩