Abstract: For a complex engineering-to-order(ETO) product, the evalua-tion results of machine hour of its parts are main basis to work out possible production program and calculate lead-time. How to measure standard machine hour and build SML for machine hour evaluation of parts are analyzed. The similarity principles and a process model for machine hour evaluation of parts are presented. In order to increase rationality and veracity of sam-ple parts selection, a SML-based identified method of similar parts is presented. Furthermore, a case-based reasoning(CBR) and SML-based evaluation method of machine hour of a part is presented. Taking machine hour evaluation of the spindle of industry stream turbine as an example, the CBR and SML-based evaluation method are verified and evaluation error is analyzed.

Key words: Mass customization, Case-based reasoning(CBR), Machine hour evaluation, Tabular layouts of article characteristics