Design and application of Kanban control system in a multi-stage, mixed-model assembly line

To achieve efficient organization in the process of mix-model production, this paper addresses the parameter definition and performance optimization of a Kanban control system in a multi-stage, mixed-model assembly line. Based on four decision variables, model replacing time, WIP, Kanban capacity and takt ratio, the model of kanban system is set up, and its dependence on environment is analyzed. Considering different circumstances and the constraint of no stockouts, methods of optimization are given. Finally, the example of Y Company is illustrated to examine the validity of proposed methods and to show the amount of WIP reduced drastically.