Designing systems for adaptability by means of architecture options

  • Authors:
  • Avner Engel;Tyson R. Browning

  • Affiliations:
  • Advanced Systems & Software Engineering Technology (ASSET), Israel Aerospace Industries, Ben-Gurion International Airport, Israel;Neeley School of Business, Texas Christian University, TCU Box 298530, Fort Worth, TX 76129

  • Venue:
  • Systems Engineering
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The value of a system usually diminishes over its lifetime, but some systems depreciate more slowly than others. Diminished value is due partly to the increasing needs and wants of the system's stakeholders and partly to its decreasing capabilities relative to emerging alternatives. Thus, systems are replaced or upgraded at substantial cost and disruption. If a system is designed to be changed and upgraded easily, however, this adaptability may increase its lifetime value. How can adaptability be designed into a system so that it will provide increased value over its lifetime? This paper describes the problem and an approach to its mitigation, adopting the concept of real options from the field of economics, extending it to the field of systems architecture, and coining the term architecture options for this next-generation method and the associated tools for design for adaptability. Architecture options provide a quantitative means of optimizing a system architecture to maximize its lifetime value. This paper provides two quantitative models to assess the value of architecture adaptability. First, we define three metrics—component adaptability factors, component option values, and interface cost factors—which are used in a static model to evaluate architecture adaptability during the design of new systems. Second, we enhance a dynamic model to evaluate architecture adaptability over the maintenance and upgrade lifetime of a system, formulating a Design for Dynamic Value (DDV) optimization model. We illustrate both models with quantitative examples and also discuss how to obtain the socio-economic data required for each model. © 2008 Wiley Periodicals, Inc. Syst Eng