Project information

Title: Aspect-Oriented, Model-Driven Product Line Engineering
Start date: 01-10-2006
End date: 30-09-2009

Project description

The focus of the project, which is coordinated by Lancaster University, is on combining aspect-oriented and model-driven techniques to address variability at each stage in a software product line engineering lifecycle and bind the variation points in various development stages and dimensions into a coherent variability framework across the life cycle.

The specific focus of this research position will be the modeling of the design process for product line architectures, where traceability issues will be taken into account explicitly, in all phases, from requirements, via product line architectures, to (UML) design and implementation.

This requires at least two important extensions to the current approaches:

  1. Software development process must be modeled in such a way that all the information necessary for tracing can be captured and represented adequately;
  2. When necessary, the artifacts used in models must be extended in such a way that the required information can be recorded and linked.

In particular, tracing aspects is in important requirement for both in modeling processes and artifacts in models. During the software development process, many decisions are taken among -generally- a large set of alternatives. In order to trace the decisions and if necessary to reengineer certain models in the changing context, not only results of the decisions but also the rationale of decisions must be modeled. Since product line approaches largely depend on technology and market forecasts, they do not always based on hard facts.

Moreover, especially in the early phases of software development, the software engineers must make many assumptions which may be justified only in a later stage of software development. These uncertainties, therefore, must be modeled.

It is very important that from the start of the modeling process the handling of uncertainty is made explicit, where all origins of uncertainty, such as imprecise requirements from stakeholders, uncertainty about future scenarios, uncertainty of the impact of design decisions on the quality of the final system, are taken into account. As a consequence, a study of appropriate techniques to represent uncertainty is envisaged.

Participants from TRESE

Some publications

  • Elimination of Constraints from Feature Trees (Pim van den Broek, Ismenia Galvao, Joost Noppen) Workshop on Analysis of Software Product Lines (ASPL 2008) in Proceedings of the 12th International Software Product Line Conference (SPLC 2008), 2008
  • Traceability for Model Driven, Software Product Line Engineering (Nicolas Anquetil, Birgit Grammel, Ismenia Galvao, Joost Noppen, Safoora Shakil Khan, Hugo Arboleda, Awais Rashid, Alessandro Garcia) ECMDA Traceability Workshop Proceedings, 2008
  • Survey of Traceability Approaches in Model Driven Engineering (Ismenia Galvao, Arda Goknil) the Eleventh International IEEE EDOC Conference (EDOC 2007), 2007