Abstract

Cellular manufacturing (CM) has emerged because of the need for manufacturing organizations to supply products that require more customization and that have shorter product life cycles and times to market. This shift from mass production to demands for mid-volume and mid-variety product mixes has shown that traditional manufacturing systems relying on functional or line layouts are not efficient and flexible enough. CM is an alternate manufacturing system combining the high throughput rates of line layouts with the flexibility offered by functional layouts (job shops). The benefits include reduced set-up times, material handling, in-process inventory, better product quality and market response time. The benefits of CM can only be achieved by sufficiently incorporating the real-life structural and operational features of a manufacturing plant when creating the cellular layout. This research presents two integrated CM models, with an extensive coverage of important manufacturing structural and operational features. The first CM model is initially formulated as a mixed integer non-linear program that incorporates multi-period production planning and dynamic system reconfiguration with deterministic production requirements. It consists of alternate process routings, the operation sequence of parts, machine duplication, machine procurement and machine capacity. Preliminary computational experiments show that only small-scale problems can be solved using the developed mixed integer non-linear program. This warrants the linearization procedures that are proposed to convert it into a linearized mixed integer programming formulation so as to solve problems of larger scale. This linearized mixed integer program is first solved using an exact solution (ES) procedure through the simplex-based branch and cut procedure of CPLEX software. Computational results are presented by solving some small-scale to large-scale numerical examples, extracted from existing literature. Although small to medium-sized problems can be solved using ES, larger ones (representing real-life sized problems) cannot be solved within reasonable computational times. A tabu search meta-heuristic is, therefore, developed to solve this mixed integer linear program. The results show that the tabu search procedure generates good quality solutions for real-life size problems within acceptable computational times. The second model addresses the same attributes as the first one but an important extension is the introduction of routing flexibility in the system by the formation of additional alternate part process routings, called contingency routings . In addition to the alternate main process routings being created for each part type, the model forms contingency routings , which serve as backups so as to effectively address the reality of part process routing disruptions, thus allowing the cellular manufacturing system to operate in a continuous manner even in the event of any breakdowns. This enhanced model is formulated as a mixed integer linear program and is then solved using the simplex-based branch and cut procedure of CPLEX (ES). The results show that the contingency and main routings can be formed simultaneously, thus bringing enhanced system and routing flexibility. The trade-off between the increased system flexibility obtained versus the additional cost to be incurred through the formation of contingency routings for all parts is discussed by comparing the solutions obtained from the model without routing flexibility and the one with routing flexibility. It is found that the enhanced system flexibility enabled by contingency routings offsets the additional system investment costs. Keywords : integrated cellular manufacturing systems design, mixed integer programming, linearization procedures, simplex-based branch and cut procedure, tabu search, routing flexibility, contingency routings, enhanced CM system flexibility.