Hidden Markov model-based autonomous manufacturing task orchestration in smart shop floors

Smart manufacturing requires flexible production organization and management to handle the dynamic customer requirements rapidly and efficiently. In the context of smart manufacturing, work-in-progress (WIP), machines, and other physical resources in smart shop floors are endowed with intelligence, such as self-perception and self-decision-making. In this situation, the manufacturing task orchestration in such smart shop floors becomes autonomous, which is different from the traditional one that is centrally set and managed. The manufacturing tasks are accomplished with the help of autonomous communication between the WIP and the machines. This paper firstly clarifies the logic of autonomous manufacturing, in which the core idea is the autonomous communication and collaboration between the WIP and the machines during production. Furthermore, the autonomous manufacturing task orchestration (AMTO) problem is described. An improved hidden Markov model (HMM) is proposed to formulate the problem and generate an optimal AMTO solution for a certain process flow. A demonstrative case is implemented to verify the feasibility of the proposed model and method. The results show that HMM can give suggestions on AMTO and dynamically adjust the situation based on the real-time manufacturing data.