Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Industrial Engineering


Cho, Byung R

Committee Member

Kurz , Mary E

Committee Member

Taaffe , Kevin

Committee Member

Williams , Calvin L


In this dissertation, we intend to integrate censoring into robust-tolerance engineering using improved dual response surface modeling. This is perhaps the first attempt in the quality control literature. In the literature, response surface models have largely been restricted to second order (or quadratic) models and robust-tolerance designs have been restricted to situations involving complete observations. We intend to show that higher order response surface models can be more powerful in terms of prediction ability, and are therefore more reliable than the preferred quadratic models in the general context of robust design. We will also consider formulating robust and tolerance designs in the presence of censored data. This is especially important for lifetime studies, where experiments are designed to determine the expected lifetimes of products under a variety of conditions. It is most often necessary to terminate experiments of this nature before the failure of the all the elements in the sample. Thus, lifetimes are observed for failed items, but censored times are observed for surviving items only. Available robust design methodologies in the literature have paid very little or no attention to problems of this nature. The proposed study is naturally suited for larger-the-better type (L-type) quality characteristics. As a result to this, we intend to propose quality loss functions that properly model such quality characteristics in a very intuitive and practical way. At the conclusion of the study, we intend to develop a censored-robust tolerance design optimization procedure, which will integrate all the major concepts of this dissertation.