Production planning and scheduling in a flexible manufacturing system environment

In this dissertation research, two decision support models are formulated in which the production scheduling function in a flexible manufacturing system is addressed. Specifically, the decisions addressed by the models include the break-up of weekly production requirements to daily production requirements, defining the leakage between MRP lot sizing and shop order release, the determination of production batch sizes for each operation of each part-type, the determination of inventory and backordering levels, and the daily assignment of each batch to machine groups, subject to the flexibility of alternate routings.;Both the models are formulated as multiproduct, multiperiod, multistage models having an underlying network structure. The large-scale nature of the models, along with the presence of bundle capacity constraints defies the straight forward application of any standard solution code. Furthermore, the second of the two models has integer variables. A price-directive decomposition procedure using column generation is used, whereby the overall problem is decomposed into a master problem and several subproblems. The subproblems are shown to have a pure network structure and are efficiently solved using a standard network code. However, since the second formulation is an integer network model, the problem is first linearized by relaxing the setup variables. The decomposition technique is then applied to this relaxed version as part of a heuristic procedure to solve the overall integer model.;Experimentation with varying problem sizes is performed on both models to first determine the computational efficiency of the solution techniques; and second to determine the impact of routing flexibility on factors including total cost, inventory levels, existence of bottlenecks, capacity utilization, throughput time, and the number of occurrences of setups and split lots. Experimental results indicate that while the total system cost, inventory levels, number of bottleneck machines, and throughput times decrease as routing flexibility increases, there is a concurrent increase in the number of setups and split lot production. This clearly indicates that while increase in routing flexibility is desirable for some key shop performance measures, the benefits are not attained without associated costs.