COMPASS - Comprehensive Modelling for Advanced Systems of Systems

From October 2011 to September 2014
Project Leader(s): Dr. John Fitzgerald
Staff: Dr. Steve Riddle, Prof. Sascha Romanovsky, Maria Arias Martin (Research Project Administrator)
Contact: Dr. John Fitzgerald
Homepage: http://www.compass-research.eu/

The COMPASS consortium is a group of researchers and companies committed to collaborative research on model-based techniques for developing and maintaining Systems of Systems (SoS).

Modern networking technologies let systems cooperate by sharing resources and offering services to one another so that the resulting system of systems has a behaviour that is greater than just the sum of its parts. For example, the information systems of fire, police and hospital services can together offer a flexible and responsive SoS for emergency management, even though the individual systems were not intended for collaboration. At a different scale, the integration of systems on board an aircraft can offer more energy-efficient and robust flight control.

Although there are great opportunities here, the design of innovative products and services that take advantage of Systems of Systems (SoS) technology is hampered by the complexity caused by the heterogeneity and independence of the constituent systems, and the difficulty of communication between their diverse stakeholders. Developers lack models and tools to help make trade-off decisions during design and evolution leading to sub-optimal design and rework during integration and in service. Our work is inspired by the vision that complex SoSs can be successfully and cost-effectively engineered using methods and tools that promote the construction and early analysis of models.

Our research agenda involves:

Developing a modelling framework for SoS architectures.

Providing a sound, formal semantic foundation to support analysis of global SoS properties.

Building an open, extendible tools platform with integrated prototype plug-ins for model construction, simulation, test automation, static analysis by model-checking, and proof, and links to an established architectural modelling language.

Evaluating technical practice and advanced methods through substantial case studies.