SysML Modeling Mistakes and Their Impacts on Requirements

Abstract

The Systems Modeling Language (SysML) represents a significant and increasing segment of industrial support for building critical systems. Because modeling is a human-centric activity, mistakes are unavoidable. Although there exist several software defect classifications, little is known about the mistakes pertaining to SysML modeling and the implications of those mistakes in model-driven requirements engineering. In this paper, we report a systematic mapping through which 42 SysML modeling mistakes are identified from 19 primary studies. With an emphasis on the evidence of industrial relevance, we further uncover that, despite some mistakes hurt requirements satisfaction, others help make the requirements more complete and the specifications more precise. Our work sheds light on understanding the scope of the SysML mistakes and checking requirements fulfillment in the face of the mistakes.