Scheduling Supply Chains With Batchwise Fabric Dyeing Operations

Abstract

Meeting customer due dates has become important for textile coloration firms' long-term survival due to rapidly changing business conditions and intense global competition. In this dissertation, optimization of the fiber-textile-apparel-retail chain, including batchwise fabric dyeing operations, was pursued.

The performance of the Virtual Factory (VF), a job shop scheduling system developed at North Carolina State University, was tested in multi-factory, rolling horizon settings to more accurately predict how it would perform in industry by eliminating transient effects presented in previous experimentation. The VF performed well in all multi-factory supply chain environments.

By taking the theoretical approach of color physics, setup matrices for dyeing operations were developed to include four indices; fabric/dye type, hue, lightness and chroma. After refining the matrices to capture the interdependency of the hue, lightness, and chroma of colors dyed within the same fabric/dye type, an existing sequence dependent scheduling algorithm was modified accordingly. The proposed algorithm and additional modifications were implemented in the VF, and a variety of one machine flowshop scenarios were tested against another algorithm found in the scheduling literature and a lower bound approximation. The proposed algorithm performed well in tight due date ranges and with a large number of jobs but not as well under other conditions. Sensitivity analysis of the initial parameters used in the proposed algorithm showed that its performance is highly dependent on these values.