DEA advanced models for geometric evaluation of used lathes

Abstract

This paper presents a preliminary study of the quality of lathes using Data Envelopment Analysis (DEA) as a multicriteria tool. Its purpose is to aggregate several measurements that show geometrical errors to obtain one single measurement that will be free of subjective opinions. Due to the small number of DMUs, additional techniques other than the DEA CCR model (super efficiency, cross-evaluation, inverted frontier and weights restrictions) are reviewed and used to differentiate between otherwise identical efficiency scores. The specificity of this problem required some changes in the above-mentioned methods. The models were applied to four lathes at the Fluminense Federal University Machine Shop.