This experiment will lead to a better understanding of the potential of formal methods to improve quality and reliability of safety critical embedded systems. In particular, GSE is currently involved in the development of an Aero Engine Electronic Control (EEC) System Test Facility (STF) in conjunction with BMW Rolls-Royce GmbH (BRR) and we propose to implement part of the EEC STF using both traditional and formal software development methodologies.
The baseline project using traditional methods will be completed at the commencement of this application experiment. Therefore the application domain will be well understood and data concerning the traditional development process will be available for comparison with the experiment's results.
By re-implementing the same piece of software using now more rigid methodologies, namely the RAISE method and tools, we expect to develop insights into the use of formal methods, how these methods compare to traditional methods, and whether these methods lead to improvements in efficient, software quality, or reliability as claimed by their proponents. This latter question is especially important in the context of safety critical systems such as the EEC.