A systematic approach to the design of weaving preparation systems operating under variable demand

In today's marketplace, textile enterprises are being required to respond to customer demand for a wider range of products with shorter delivery times. The metric used to describe this type of customer demand has been loosely characterized as quick response or demand activated manufacturing. The end result for textile producers is a greater variability in both order mix and demand level. An area that has received limited attention is the development of methodologies to assist textile producers in deploying quick response manufacturing systems that meet customer demand while maintaining acceptable efficiency and profitability. The goal of this research is to define a framework for the design of quick response textile processing systems that is capable of selecting the optimal system in terms of capital and operating costs from a range of possible technological alternatives, while accounting for both variable demand and the stochastic nature of these processes. This is accomplished by iterating between design and assessment problems. The design screening formulation is based on overestimating system performance and underestimating system cost with deterministic product demand and processing times. The design assessment problem carries out a more rigorous characterization of performance and cost based on the probabilistic distribution of demand and process times. This framework is implemented for the design of weaving preparation systems, a critical bottleneck in the conversion of yarns into woven fabric. It will be shown that mixed-integer linear programming techniques can be used to solve the design selection problem, and discrete-event simulation is effective for the design assessment.