Cellular layout design using Tabu search, a case study

Abstract

This paper presents a facility layout design for a cellular manufacturing system (CMS), with the automated guided vehicles (AGVs) as the transportation device, to minimize the total material handling costs. A new concept of assigning the workstations to the non-overlapping closed zones, by which independent AGVs are allocated for internal transfer of materials/parts in each zone, is proposed. To handle this problem, a mathematical programming model is developed. The problem has been shown to be NP-hard. This computational difficulty has led us to consider suboptimal solutions generated by a Tabu search with three algorithmic variants. To evaluate the efficiencies of these algorithms, an ANOVA statistical test is performed and the best algorithm is designated. The solution of the new algorithms is then compared with the solution obtained by the CPLEX software. The result indicates that the designated algorithm can provide an average solution with a small deviation of only 0.01% from the optimal solution. This algorithm is also applied to a real-world case problem. The use of this algorithm for re-designing of the layout for the real-world case problem shows considerable cost saving comparing to its existing layout design.