Detection of machine failure: Hidden Markov Model approach

Hidden Markov Models (HMMs) are widely used in applied sciences and engineering. The potential applications in manufacturing industries have not yet been fully explored. In this paper, we propose to apply HMM to detect machine failure in process control. We propose models for both cases of indistinguishable production units and distinguishable production units. Numerical examples are given to illustrate the effectiveness of the proposed models.