A systematic approach towards manufacturing health management

The complex and dynamic nature of the modern manufacturing environments place an additional burden on the management of system health and maintenance of machineries. In this paper, we propose an integrated solution to fault diagnostics and prognostics that aims to facilitate system health management in a manufacturing environment. The proposed solution addresses the drawbacks of the commonly adopted practice of reactive fault diagnostics approach, in particular: (1) the reusability of the previously developed diagnostics techniques; (2) the loss of time and efforts to revisit previous issues due to lack of central management system; and (3) the lack of mechanism to perform diagnostics knowledge retention. We achieve these objectives in our system through the use of two-pronged approach: (1) the modularization of commonly used analytics algorithms; and (2) the abstraction of the analysis process. The applicability of our proposed solution is demonstrated on the fault diagnostics of a machine fault simulator. From the experimental results, we firmly believe that we have a viable system oriented solution to the underlying problems in providing diagnostics for modern manufacturing environment.