Condition monitoring and data analysis are key factors to avoid unexpected catastrophic breakdowns and detect incipient failures. Managing of these processes are not easy in complex system which has many inputs/outputs and interdependent relationships. In conventional methods, the condition monitoring is carried out separately and systematically due to the large number of monitoring point. This paper presents an intelligent condition monitoring platform (ICMP). This platform uses multi-agent system (MAS) advantages via mathematical process named failure sensitive matrix (FSM). Through this platform, the whole system is dynamically monitored, despite all the complexity. Cause and effect relations of a fault are evaluated more comprehensively and the success of achieving global results is higher. The growth of monitoring equipment or system is not a problem, because matrix structure is expandable as wanted and its performance is fast due to the nature of the parallel computation. Other important advantage of this platform is that it is an interface which works with expert systems and artificial intelligence techniques with no restrictions.