Artificial Intelligence in Enabling Predictive Maintenance and Maximizing Machine Tools Operational Life

Condition based monitoring associated with predictive maintenance are popular areas of research in the context of industry 4.0. Due to increased demand and competitive market, it is very important for the companies and therefore their production lines to be able to operate reliably. Artificial intelligence (AI) has played a key role in enabling maintenance due for the machine tools. But the existing work is still far from being mature. Among the available opportunities for the improvement in the predictive maintenance of machine tools using AI after large data acquisition, is the incorporation of uncertainty in the decision-making involved. When it comes to decision-making in the presence of uncertainty, the most popular method used in the AI literature is Markov Decision Process (MDP). MDP models an AI problem as five components, namely, states, decisions, transition probabilities, cost function, and a discount factor. Once an MDP model is developed, the optimal decision-making policy is computed using stochastic dynamic programming. Therefore, we propose to develop an MDP model for predictive maintenance and health management of machine tools.