Conceptually, web-integration predictive maintenance approaches are potential to meet the higher requirements of global equipment maintenance, such as zero-downtime and minimized maintenance cost. However, due to the complexity of real global manufacturing environment and limitations of predictive maintenance technology and information technology, it is a big challenge to set up a platform to implement a web-enabled predictive maintenance system.

In this paper, an agent-based platform is presented for web-enabled equipment predictive maintenance system. The basic agent structure of a web-enabled predictive maintenance system is defined. Based on the definition, the structure of web-enabled predictive maintenance system for single machine is analyzed. It is then extended to the factory level with many different equipments and the enterprise level for global production.