IIoT-Based Prognostic Health Management Using a Markov Decision Process Approach

Recent advances in Industrial Internet of Things (IIoT) made them a key component of the Industry 4.0. Thus, several aspects of the latter, such as scheduling maintenance operations, could benefit from the existing IIoT infrastructure. We consider an IIoT-based Prognostic Health Management network for industrial facilities. Our objective is to characterize the optimal maintenance policy that favors grouping maintenance operations while reducing the deterioration and failure costs. We rely on Markov Decision Process with full information Theory to develop a realistic model for the IIoT-based PHM system in an industrial facility with multiple components prone to failure. We investigate the structural properties of optimal policies and provide numerical investigations.