A Multi-Agent based distributed fault diagnosis reference model is important for high speed and automation. Fault diagnosis and maintenance are vital aspects in industrial process, in this sense, fault diagnosis systems should support decision-making tools, new diagnosis approaches and techniques, updating of the devices. In this paper a Multi-Agent based fault diagnosis reference model is presented which combines the existing models and Multi-Agent. This model is based on a generic framework using Multi-Agent System; in this sense, the fault diagnosis problem is viewed like a feedback control process and the actions are related to the decision-making in the scheduling of the preventive maintenance task and the running of preventive and corrective specific maintenance tasks. As a result, a particular prototype design based on the JADE platform is obtained. This new model is compared to some important existing models and applied to a real investigation. This Multi-Agent based model provides a comprehensive treatment of the fault management problem, it can be used in any level of industrial environments and it can be integrated with the supervision applications without modifying the current automation architecture.