Exploratory Assembly Process Evaluation through Manufacturing Simulation

Abstract

The rise of mass customization imposes significant challenges for manufacturing companies regarding management and development of manufacturing systems, where advanced digital technologies such as simulation can serve as central enablers. The assembly process of the truck manufacturer Volvo Trucks in Tuve handles a broad range of customized product variants. This imposes complexity in process redesign and rebalancing, increases reliance on buffers and challenges efficient resource utilization. In this study, Discrete Event Simulation is applied as a data-driven and visual support tool for iterative, Lean oriented redesign and evaluation of a manual assembly process, supporting close collaboration with process and data domain experts of the company. An approach for handling of detailed Predetermined Time System process data and production levelling rules is applied, generating a detailed, stochastic simulation of product based variation, enabling the model user to conduct subsequent experiments covering changes in both demand and levelling restrictions. The developed model is used as a basis for analyzing the effects of implementation of parallelized, stationary assembly and kitting, where the resulting workload and manning requirements, as well as required buffer levels safety margins in a Kanban implementation, are evaluated with regards to product based variation and production pace. The study indicates the usefulness of Discrete Event Simulation in mass customization environments, and underlines challenges of data management and model complexity. The study identifies a number of benefits in the implementation of kitting and parallelization, as well as several drawbacks and potential risks.