24th CIRP Design Conference A conceptual basis for inconsistency management in Model-Based Systems Engineering

A crucial issue in system architecting is the need to study systems from different viewpoints. These viewpoints are defined by a variety of factors, including the concerns of interest, level of abstraction, observers and context. Views conforming to these viewpoints are highly interrelated due to the concerns addressed overlapping. These interrelations and overlaps can lead to inconsistencies. The challenge is to identify and resolve - that is, manage - such inconsistencies. This paper introduces an approach to managing inconsistencies within the context of Model-Based Systems Engineering (MBSE). In current practice, the management of inconsistencies relies on ad-hoc methods and infrequently conducted activities such as reviews. The result of this practice is that decisions are often made based on inconsistent information, which can lead to costly rework or even mission failure. Therefore, assisting humans by means of a computational method that can continuously identify and aid in resolving inconsistencies adds significant value. In the paper, the hypothesis that pattern matching can serve as a generic means of identifying inconsistencies is investigated. It is shown that graph patterns can be used as a means to capture conditions for and formally reason about the existence of inconsistencies, and to specify resolution alternatives. The paper concludes that using patterns to manage inconsistencies can be very effective and accurate, but it may also incur additional costs that must be carefully balanced with the benefits gained. c