A decision analysis view of the effect of robot repeatability on profit

Abstract

The purpose of this paper is to view recent developments in non-von Neumann and Morgenstern utility functions (vNM) via a stationary statistical-economic model. Such a model is suggested in the literature for evaluating robot precision, assuming that repeatability follows a Rayleigh distribution. Various vNM and non-vNM utility forms are discussed and illustrated for the robot evaluation and selection problem. The approach suggested here may be applicable for companies operating under profit centers where market systematic risk data are not available and decisions are based on the decision maker's (DM) attitude toward risk. Although the numerical illustrations demonstrate a consistency among the expected profit measure for most utility forms that were investigated, the analysis provides an extended framework for dealing with risky production problems. Furthermore, it illustrates the sensitivity of robot operational decisions such as speed and of both the output rate per time unit and the product quality to the DM's attitude toward risk. It allows the DM to analyze and explain decisions as affected by the selection of the utility form, as well as the technological and economic parameters and measures describing the problem, particularly the technical risk involved in the determination of a robot operational problem.