Using DEA to optimize a mixture design with multiple responses incorporating cost considerations

Mixture design (MD), which is commonly adopted in chemical-manufacturing, food-processing or pharmaceutical industry, is more suitable than other design of experiment methods when component proportion is a crucial influence on product quality. This study develops an optimization procedure, which utilizes Data envelopment analysis to analyze multiple inputs/outputs and costs for experiments with mixtures. Proposed procedure can optimize MD with multiple responses while maintaining low manufacturing costs, and thus enhance manufacturer competitive advantage. As demonstrated in the case of formulation design for automobile rubber bowls, proposed procedure optimizes experiments with mixtures while minimizing costs.