Abstract
A novel production line with high automation, flexibility, reliability and reconfigurability, which is designed for the smart factory and named as automated flexible transfer line (AFTL), is studied in this paper. Different from the other production lines, AFTL consists of sequential and reconfigurable stages which are grouped by multiple machining cells, and each machining cell is composed of a single robot and several machines. This special line structure contributes to the complexity of the non-linear line cycle time relations, and the cost of the line contains the cost of machines, robots and stages. The aim is to balance and configure the AFTL in minimal cost with a given line cycle time, which is equivalent to find the appropriate subsets of operations and assign each of them to the stage with an optimal configuration. Three novel and efficient lower bounds in different levels based on solving the set partitioning problem in AFTL design problem are presented and an effective algorithm is developed. The experimental results and case problem results prove that the proposed algorithm together with the lower bounds are effective and applicable for the industrial cases.