Advanced testing and evolution of executable behavioural software models

Description

Model-driven software engineering (MDSE) relies on the systematic use of (often visual) software models as primary artefacts throughout the software development lifecycle. Modelling languages based on statecharts and Petri-net allow to model the executable behavior of a system. Ensuring and maintaining the quality and reliability of such behavioural models over time, finding scalable solutions to manage a multitude of interacting models, and coping with their (co-evolution) remains very challenging. The research aims to address these challenges by studying, combining and applying wellestablished software development techniques at the modeling level. The goal is to increase the state-of-the-art along the following directions:

• Advanced support for testing and lightweight formal verification of executable behavioural models, as well as for analysing and improving model quality.

• Scalable solutions for composing multiple interacting behavioural models, supporting model co-evolution, inconsistency management, design space exploration and variability analysis.