Product conceptual design based on theory of belief functions

In the development phase, the designer makes decisions under uncertainty, contradiction and ignorance conditions. Are these decisions correct and to what extent? How much do they influence customer's satisfaction? These are only some of the questions the designers face all time. These dilemmas appear more in early phases of the development process. The explicit objectives of this paper is improvement of design decision making process in such a way that in same time when designer made a decision he or she show result that decision on customer satisfaction. In order to solve above problems, the paper describes the method, which enables integrated use of the axiomatic approach to designing, and the Taguchi method of robust design. This approach implies the modelling of the development process as an evidence-reasoning network based on uncertain evidence described via belief functions of (Dempster - Shafer theory). The paper starts with base concept of belief functions and valuation based system or evidential reasoning system for representation and reasoning based on uncertainty. After that we introduce coefficient of relative decrease of uncertainty in design process. On the base of above we gave a new graphical representation of system architecture - the evidence networks and represent first and second design axiom in uncertainty condition. On the end through example of design belt-flywheel we presents new hybrid paradigm for product/manufacturing design in uncertainty condition.