AuthorsR. Behjati, T. Yue, L. C. Briand and B. Selic
TitleSimPL: a Product-Line Modeling Methodology for Families of Integrated Control Systems
AfilliationSoftware Engineering, The Certus Centre (SFI), Software Engineering
Project(s)The Certus Centre (SFI)
StatusPublished
Publication TypeJournal Article
Year of Publication2013
JournalInformation and Software Technology
Volume55
Number3
Pagination607-629
Publisher.
Abstract

Context: Integrated control systems (ICSs) are heterogeneous systems where software and hardware components are integrated to control and monitor physical devices and processes. A family of ICSs share the same software code base, which is configured differently for each product to form a unique installation. Due to the complexity of ICSs and inadequate automation support, product configuration in this context is typically error-prone and costly. Objective: As a first step to overcome these challenges, we propose a UML-based product-line modeling methodology that provides a foundation for semi-automated product configuration in the specific context of ICSs. Method: We performed a comprehensive domain analysis to identify characteristics of ICS families, and their configuration challenges. Based on this, we formulated the characteristics of an adequate configuration solution, and derived from them a set of modeling requirements for a model-based solution to configuration. The SimPL methodology is proposed to fulfill these requirements. Results: To evaluate the ability of SimPL to fulfill the modeling requirements, we applied it to a large-scale industrial case study. Our experience with the case study shows that SimPL is adequate to provide a model of the product family that meets the modeling requirements. Further evaluation is still required to assess the applicability and scalability of SimPL in practice. Doing this requires conducting field studies with human subjects and is left for future work. Conclusion: We conclude that configuration in ICSs requires better automation support, and UML-based approaches to product family modeling can be tailored to provide the required foundation.