Coordination Strategies and Individual Behavior in Complex Engineered Systems Design

Abstract

Coordination – ensuring interfacing groups are working with consistent data, interpretations of that data, and consistent goals – becomes necessary when design work is distributed or partitioned among different individuals or teams. System design optimization algorithms are well-established for the coordination of analytical design problems. However, in actual system design, individuals are unlikely to coordinate by following algorithmic procedures. In a design organization, distributed tasks must be first partitioned so that they can be worked in parallel, and then coordinated so that the results can be joined together to effect the overall project goal. In this organizational context, coordination is primarily a communicative process focused on information sharing among parallel tasks. This research focuses on these individual communication behaviors and demonstrates their impact on system-level coordination.

This dissertation addresses two research questions. First, what approaches and behaviors do individuals use to facilitate system-level coordination of distributed design work? This is answered with a pair of exploratory studies. A qualitative study based on interviews of industry experts found that proactive, empathetic leadership-based behaviors and more passive, authority-based behaviors are complementary approaches to facilitating coordination. A quantitative study based on a survey of novice designers resulted in identification of coordination roles within teams. Text analysis and network analysis of survey data identified five roles: some are more communicative and focused on integrative tasks and leadership, whereas others are less communicative and focused on documentation and detailed engineering tasks. The findings suggest a parallel to the active and passive behaviors identified in the qualitative study.

The second research question asks what is the quantitative impact of these behaviors on system-level performance. A descriptive agent-based model was developed that simulates the impact of individual behaviors on the system-level performance of a distributed design task requiring coordination. Results from this model indicate a correlation between more active agent behaviors and higher performance, but at the cost of increased peer-to-peer interactions.

This dissertation illustrates the importance of a balance between proactive, empathetic-leadership-based and more passive, authority-based processes and behaviors for the effective coordination of decomposition-based design work. There are three primary contributions of this work. This dissertation highlights that coordination is a central task of systems engineering personnel in engineering design organizations. This work uniquely shows how theory pertaining to coordination applies to and describes systems engineering activities, particularly in relation to engineering design. The second contribution of this work is the development of quantitative approaches to describe and evaluate coordination practice in decomposition-based system design. This is shown through an agent-based simulation model to assess the impact of communicative and information-seeking behavior on coordination and design task outcomes. This model also serves as a platform that allows extensive parametric analysis, permitting exploration of a variety of design tasks, organizational structures, and agent behaviors. Finally, this work illustrates the importance of coordinator roles in effective distributed design tasks. Frequently, managerial and integrative roles are not included in measures of coordination effectiveness, which we argue is not representative of true systems engineering practice. This dissertation is a step towards the inclusion of systems engineering and management roles in models of coordination effectiveness.