Model-based Product Derivation: Applied Formal Methods in the Presence of Explicit Variability and Commonality

Funded by Science Foundation Ireland Lero CSET

The goal of this project is to enable systematic application artifact and product tracking, evolution and derivation within a product line and to improve the degree of automation within the product derivation process. The main objective of this project is to combine formal approaches to describing requirements, software architecture and other artifacts with model-based development approaches to support formal automated derivation of verifiable products within a product line.

The high-level objectives of the project are:

• Determine a variability model that is derivable from the requirements model (both domain and application) by using more formal descriptions of the domain and application requirements.
• Investigate how the variability and requirements models (both domain and application) can be used to support formal generation of the architecture models and other artifact models.
• Establish a formal approach to the definition of variability between the variability, requirements, architecture and other artifact models both in the domain engineering and application engineering within software product lines.

In order to enable more automated reasoning about these artefacts in terms of transformation, organization, search and discovery, etc. there must be a fundamental semantics to the software product line domain and the specification of artefacts is paramount. This project will define such a semantics by using evolutions of existing formal approaches to the description of these various artifacts. More specifically, a logic called kind theory be used to describe artefacts and their interrelationships and this logic will be integrated into the Cadena integrated environment for defining and modeling component-based systems in model-driven architectures.