An Integrated Assembly Line Balancing and Maintenance Model Using a Markov Decision Process Incorporating Machine Status

Abstract

This study introduces a novel assembly line balancing model that integrates operational and maintenance requirements. The objective is to minimize the operational costs, which includes costs related to tools\workers, workstations construction, and repairs. The methodology focuses on dynamically balancing tasks distribution across workstations to reduce cycle time while adhering to operational constraints; precedence and resources requirements. We employ a Markov Decision Process (MDP) framework, enabling real-time, data-driven decision-making.The MDP framework facilitates adaptive repair actions, with decisions based on real-time assessments of workstation health, tools\workers needs, and repair actions. By considering real-time machine health assessments, the model dynamically assign  tasks and allocate resources across the production line.The MDP-based model demonstrates significant improvements in both cycle time and cost efficiency, creating a more resilient production environment. By adapting to changes in workstation health and operational demands, the model ensures flexibility in responding to the challenges of modern manufacturing. This framework not only enhances cost-effectiveness but also supports sustainable production, where resources are allocated efficiently, machines operate at peak performance, and system durability is optimized. Initial experimentation using datasets provides insights into the effectiveness of our approach