Master's Defense

Sean Ryan Mueller

(Advisor: Prof. Dimitri N. Mavris)

A Holistic Modeling and Simulation Framework of Complex 
Product Development to Support Engineering Management

Friday, May 2 
9:00 a.m.
Teams (virtually)

Abstract
The development of new products by organizations presents many challenges. New complex and multidisciplinary products, increased customer demands and competitive pressure, as well as agile, collaborative, and concurrent processes, make developing products increasingly complicated and challenging. This requires extensive planning and management of the organization, processes, and resource allocations to develop products efficiently. Increasingly capable digital tools are available that can address these challenges and accelerate the development of new products by facilitating collaboration, information sharing, and knowledge creation. However, their implementation also is challenging and the results are uncertain.

Traditional methods for planning product development often rely on the experience of managers, are based on assumptions, and have little to no quantitative basis, resulting in frequent cost and schedule overruns, as well as quality problems. Modeling and simulation can help manage the product development process by enabling risk-free testing and evaluation of different what-if scenarios. Some methods for this already exist in the literature but have seen limited application in industrial practice due to the lack of a holistic view of product development. In particular, the impact of digital tools on product development has rarely been considered in previous simulation approaches.

Therefore, a framework for holistic modeling and simulation of product development is proposed. An extensive analysis of existing methods led to the selection of a baseline model which was adapted to include aspects of all important product development domains (product, process, organization, tools). The simulation model is an agent-based model that models every entity of these domains as an agent allowing for various interactions among entities and the analysis of the emergent behavior. The most important addition are agents representing tools used for engineering and supporting activities. These influence the accuracy of validation activities and the way information is shared between engineers.

The framework was applied to a notional case study to analyze its behavior, perform sensitivity analyzes, and demonstrate its capabilities. The results proved that the implemented simulation logic behaves as intended and represents product development more realistically than the current methods. The impact of the agents of digital tools was studied extensively through multiple Design of Experiments (DOEs) to demonstrate their behavior and discover various interactions.

The proposed simulation framework can enable informed decision-making to improve product development performance by testing multiple configurations and identifying causes of inefficiencies. This is achieved through a comprehensive analysis of the entire design space of the organization and by providing detailed information through measures of effectiveness (MOEs) and productivity (MOPs) at multiple levels of granularity.

Validation based on empirical data and through experts is still necessary to prove the practical applicability of the framework. In the future, the framework could be a steppingstone toward the development of Digital Twins of Organizations, by adding more detail, further improving its capabilities, and integrating real data.

Committee

• Prof. Dimitri N. Mavris – School of Aerospace Engineering (advisor)
• Prof.  Andrei G. Fedorov– School of Mechanical Engineering
• Dr. Olivia J. Pinon Fischer – School of Aerospace Engineering
• Prof. Oliver Sawodny –   Institute for System Dynamics, University of Stuttgart
• Prof. Christina Tarín – Institute for System Dynamics, University of Stuttgart