Type:

Year:

Supervisor:

Institution:

Page(s):

Abstract:

The purpose of this research is to support decisionmaking for solving design issues in the development phase of complex systems supported by numerical simulation. We conducted our studies in a multinational car manufacturer. The first part of the research was devoted to identifying the difficulties encountered in the issue resolution process, with a particular focus on decision-making issues, methods and tools. A qualitative study done with 11 experts and on 40 decision problems highlighted that the decision-makers choose from a set of process alternatives rather than artifact alternatives. The consequences of these process alternatives such as recalculating, integrating information, waiting for the technical definition of the vehicle to evolve, etc. are not explicit. We identified the lack of a rigorous framework as an opportunity for improvement. The second part was therefore to propose a framework to support design decisions. Concurrent engineering, resources constraints and project management issues have been often overlooked in the Decision Based Design literature. Attempting to bridge this gap, we designed IRDS framework. Through IRDS, we propose to make explicit the process alternatives, to gather economic data and expert forecasts in adecision model based on the prescriptive decision theory, including the maximization of the expected utility and the economic value of imperfect information. The third part of the research is related to the impact of uncertainty on the data collection process and on the overall decision outcomes. This has been done through proposing a sensitivity analysis that is performed with available data, before data gathering through elicitation process. The impacts on the decision-making process and information exchanges between stakeholders, as well as the resources consumed by the new practices we proposed have also been studied on a more superficial level. This work was in particular deployed and tested on 5 cases studies. The validation of this approach requires to collect further empirical evidence to support the hypothesis that better decisions are made on the long run. We are confident that our research will serve as a base for future studies on the design and the implementation of frameworks addressing industrial challenges.

Keywords:

decision analysis, modeling and simulation, problem-solving, automotive industry