Abstract

AbstractIndividual collaborative embedded systems (CESs) in a collaborative system group (CSG) are typically provided by different manufacturers. Variability in such systems is pivotal for deploying a CES in different CSGs and environments. Changing requirements may entail the evolution of a CES. Such changed requirements can be manifold: individual variants of a CES are updated to fix bugs, or the manufacturer changes the entire CES product line to provide new capabilities. Both types of evolution, the variant evolution and the product line evolution, may be performed in parallel. However, neither type of evolution should lead to diverging states of CES variants and the CES product line, otherwise both would be incompatible, it would not be possible to update the CES variants, and it would not be possible to reuse bug fixes of an individual variant for the entire product line. To avoid this divergence, we present an approach for co-evolving variants and product lines, thus ensuring their consistency.

Highlights

  • Configurability and variability play a pivotal role for collaborative embedded systems (CESs)

  • This is because our definition of both change operations and change conflicts is artifact-independent and we address the integration in the common product line development process

  • The concrete artifacts that are subject to change operations must be defined and an instantiation of their granularity levels must be provided. The latter is of specific importance, because the granularity plays a pivotal role in deciding whether a conflict exists or not

Read more

Summary

18.1 Introduction

Configurability and variability play a pivotal role for collaborative embedded systems (CESs). This has several advantages: first, the complexity of implementing such changes is comparably low as the impact on other variants does not have to be considered; second, the time required to deploy new changes and the costs are low as well This procedure is interesting for variability of CESs. Typically, a CES is used in multiple different CSGs by different companies. This is important to allow these changes to be propagated to product line level This task is challenging as, typically, the information about which part of a variant stems from which feature is not preserved when a variant is derived. The variant changes are transferred semi-automatically to the respective product line at DE level To this end, regression deltas between original artifacts and modified artifacts are computed and mapped to the respective feature at DE level. Product line artifacts are updated with the most recent changes at the AE level without the need for additional costs to redevelop the variant changes for the entire product line

18.2 Product Line Engineering
18.3.1 The Challenge of Propagating Updates
18.3.2 Artifact Evolution and Co-Changes
18.3.3 Changes to the Variant Derivation Process
18.3.4 Applicability and Limitations
18.3.5 Implementation
18.4.1 The Challenge of Lifting Changes
18.4.2 A Process for Lifting Changes
18.4.3 Deducing Feature Information
18.4.4 Applicability and Limitations
18.5 Conclusion
18.6 Literature
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call