Investment evaluation of automotive body assembly system alternative.

Abstract

The decision to invest in an automotive body assembly system is one of the most important and difficult tasks faced by auto companies because of the financial risk of the high investment and uncertainty relating to future demand. Moreover, the absence of an integrated approach for systematically evaluating system alternatives also makes this decision difficult. This dissertation aims to develop an investment decision analysis methodology and decision support tools for evaluating alternative automotive body assembly systems. This methodology and set of decision tools allow a decision maker not only to analyze and integrate various parameters in the investment analysis, but also to incorporate uncertainties associated with these parameters. Our approach consists of a series of three quantitative analyses: life-cycle cost (LCC) analysis, sensitivity analysis of the LCC model, and analytic hierarchy process (AHP). This dissertation presents the LCC analysis as a key economic evaluation of system alternatives. We have developed a deterministic LCC model and software for calculating the net present costs (NPCs)---the net present values of LCCs---of system alternatives. This is because automated systems often provide long-term benefits, such as flexibility; therefore, an economic criterion (i.e., the NPC) should be capable of capturing these benefits. We also perform a sensitivity analysis using the statistical design of experiments for investigating the effects of changes in various LCC parameters on the NPCs of system alternatives. The novel application of the regression models, developed from sensitivity analysis, derives more precise information from the LCC model. In addition, we also apply the AHP for the evaluation of integrated economic and non-economic criteria in the investment analysis. This is because automated systems may provide other intangible benefits; therefore, non-economic criteria measuring these benefits should be evaluated in the investment analysis. Both deterministic (single estimated values of input data) and probabilistic (probability distributions of input data) AHP are presented. Uncertainties associated with estimated input data (internal uncertainty) and multiple decision scenarios (external uncertainty) are also incorporated in the AHP. To illustrate the implementation and benefits of the proposed approach, we present a case example. The results of the case example represent accurate and realistic information as required by the decision maker. The decision maker can feel more confident because each system alternative is described in great detail due to the consideration of multiple criteria and scenarios and the incorporation of parameter uncertainties in our investment analysis.