Software developers are incessantly inundated with wave upon wave of offered solutions to their many pains (pains that are unfortunately and ultimately felt by their clients): third-generation languages, object-oriented programming, CASE tools, aspects, components, programming frameworks, extreme programming, and agile methods, to name just a few. Some of these solutions have indeed dramatically impacted how software is developed, while others have proved mere passing fads, never fulfilling the potential claimed for them. One of the more recent entries in this multitude is model-based software engineering (MBSE). From its emergence in the 1990s, this approach to software development along with its accompanying technologies have been promoted by advocates as game changers, promising quantum leaps in productivity and product quality. Following the initial excitement and hype generated around MBSE, its position in the limelight is now slowly fading, displaced by more recent cure-alls. While MBSE is being used in some enterprises, it is far from being the dominant software development paradigm that its proponents had hoped for. For many software professionals, its relevance and impact are unclear at best.
In this talk, Bran will first examine the essential precepts of MBSE and the value proposition claimed for it. Next, in order to understand the reality behind it – as opposed to the hype -- we will review the current industry experience with MBSE, based on thorough survey of published data. We conclude with a critical assessment of the real impact that MBSE has had to date, and what the future might hold for it. ·