An MDP-based Method for Dynamic Workforce Allocation in Bernoulli Serial Production Lines

With the development of Industry 4.0 in manufacturing, new technologies such as big data analytics, artificial intelligence, and cloud computing, have been widely deployed to production practice for performance evaluation and analysis. These technologies enabled fast and accurate production decision-making on the factory floor. To further support these activities, it's essential to develop real-time automated decision-making mechanisms backed by rigorous control and optimization analytics. The focus of this paper is on enhancing the throughput of production processes through the control and optimization of the floating workforce resources on the factory floor. Specifically, we consider serial production lines with finite buffers and machines characterized by the Bernoulli reliability model. Additionally, we assume that multiple shared workforce units can be allocated dynamically during production via real-time production bottleneck identification and mitigation in order to improve steady-state throughput. This paper addresses this problem for three and four-machine line cases, where two shared workforce units are dynamically allocated among the machines. An optimal control policy is derived based on the Markov decision process (MDP) approach. Numerical experiments are carried out to demonstrate the efficacy of the proposed method.