Where has all the Engineering gone? And how we might put it back: an Investigation into the Integration of a Systems Engineering Framework for the Development of System Dynamics Models

Abstract

System dynamics allows us to discover and explore the dynamics behind the complex problems commonly seen in management scenarios that defy our best efforts to resolve them. Yet despite all it has to offer, system dynamics modelling has been less than universally accepted as a management tool. System dynamics has its origins in control theory where robust design and calculation are fundamentals, and in posing the question, “Where has all the engineering gone?” this thesis traces the evolution of system dynamics away from its engineering roots. It explores the increasing trend in system dynamics interventions to portray the conceptual causal structure of complex problems through qualitative mapping and then to rely on the conceptualisation and largely untested assumptions as the basis for creating computational models. This deviation, in conjunction with object oriented software packages, has increased the accessibility of system dynamics widely. The risk however is that that the consequent quantitative analysis can lead to erroneous and misleading conclusions, discrediting the approach.

To build rigorous, insightful system dynamics models demands diligent application of a robust methodology. In addressing this need, this thesis seeks to answer the question “Can systems engineering put the engineering back?” The utility of applying systems engineering principles to put the engineering back into system dynamics to support credible and transparent design outcomes is investigated. Through the detailed comparison and analysis of design philosophies, the theoretical basis for the application of systems engineering to system dynamics is established. The principles of method, which form the basis of an integrated, design focused approach to system dynamics modelling are developed and then presented through application to a case tutorial. The resultant modelling project show that the clarity, documentation, and traceability of models back to the modelling requirements were increased. Difficulties in maintaining momentum in developing system dynamics analysis and stop-start engagement with the client group all threaten the quality of a system dynamics modelling intervention and the likelihood of successful implementation of the strategy.