System Interaction Complexity Metrics and Its Application to Embedded Software Systems

 

MIT SDM Systems Thinking Webinar Series

Qi Van Eikema Hommes, PhD, Research Scientist, MIT Engineering Systems Division

Date: February 28, 2011


Download presentation slides

About the Presentation

System designers have long practiced decomposition and modularization to manage the complexity of systems. The byproduct of modularization is system interactions. Complex system interactions often took designers by surprise, causing schedule delays, rework, resource overruns, system integration challenges, failures, and accidents.

This research work compared a number of system interaction complexity measures, and identified the Whitney Index (WI) and Change Cost (CC) as a complementary pair of metrics that can characterize the system interaction complexity. The in- and out- Degree Centrality help identify the causes for high WI values. The Betweenness Centrality helps identify elements that cause high CC values. Other system interaction complexities written in some of the publications were actually not very helpful (Closeness centrality, overall network centrality measures, Singular Value Index, and Visibility-Dependency Plots).

The effectiveness of WI, CC, and Centrality measures were tested and confirmed using two large, complex, embedded software systems in an actual industry setting, as well as a plastic part design case in the industry. The case study on an embedded software system for the first time quantitatively demonstrated that embedded software systems may inherently have a much higher level of system interaction complexity than that of IT software systems. The high complexity could be attributed to degradation of the architecture over time, the inherent nature of the embedded software system, and the economic incentive of software development in a closed private company environment.

About the Speaker

Dr. Van Eikema Hommes’ research work focuses on developing methodologies that improve the design and development of large complex engineered products and systems. Her recent efforts include:

  • Predicting and managing system interaction complexity in the early phase of the product design and development process
  • Better translation of customer experiential needs to engineering design specifications
  • Real options valuation for architecture flexibility in product design under market uncertainty
  • System theoretical product failure investigation process

Prior to coming to MIT, Dr. Van Eikema Hommes was a senior research scientist with the General Motors Research and Development Division. Her work there involved qualitative and quantitative market research to understand customer needs, purchase intent, and to forecast price and market share of GM’s products. She led a project using Real Options techniques to quantify the value of flexibility in engineering architecture decisions for new vehicle features in one of GM’s global vehicle platform.

In her earlier career, Dr. Van Eikema Hommes was a Powertrain system engineer at Ford Motor Company. She rotated through a number of positions from the early stage of the product development process to manufacturing, including model-based systems engineering for embedded engine control software design, requirements engineering for the Powertrain controls system, and calibration, design, and release of the exhaust subsystem, engine manufacturing quality, and lead engineer for warranty issue resolution.

Dr. Van Eikema Hommes holds MS and PhD degrees from the Mechanical Engineering Department at MIT. She is a certified Six Sigma Black Belt. She has published a number of peer-reviewed papers in the ASME DETC DTM conferences and the INCOSE conference. She authored several internal technical publications at both Ford and GM as well.

The MIT System Design and Management Program Webinar Series on Systems Thinking for Contemporary Challenges features research conducted by SDM faculty, alumni, students, and industry partners. The series is designed to disseminate information on how to employ systems thinking to address the engineering, management, and socio-political components of complex challenges.