When developing new products many different customer needs must be fulfille
d. Often a few of the needs are prioritized (as emphasized in Quality Funct
ion Deployment) and most effort in the project is focused on these needs. H
owever, even though there may be only four or five customer needs that are
important, it is not easy to develop a product that complies with all of th
ese. A well-known approach to this problem is to use Design of Experiments
to find the settings of the system parameters. In a situation with several
system characteristics, the analysis of the experiment is not as uncomplica
ted as in the case with just one system characteristic. This problem has be
en dealt with in the literature but not as much as one would expect. Tn thi
s paper we present a methodology to deal with multi-characteristic system i
mprovement, using active expert involvement. Our methodology, which is base
d on methods from four areas: Expert Groups, Design of Experiments, Conjoin
t Analysis and Optimization, is introduced using an example. Copyright (C)
2000 John Wiley & Sons, Ltd.