[Context & motivation] Requirements for automotive software systems are predominately documented in natural language and often serve as a basis for the following development process. Therefore, requirements artifact quality is important. Requirements often contain references to specific states of a system, which we call modes (e.g., “While the system is running,...”). [Problem] However, these references are often implicit and therefore, we suspect them as possible source for misunderstandings and ambiguities. [Principal idea] In this paper, we explore the relation between quality defects of natural language requirements and the description of modes within them. For this purpose, we investigate review findings of industrial requirements specifications and assess howmany findings contain issues addressing a mode and which defect types are most affected by mode-related findings. [Contribution] Our preliminary results showthat 46% of all considered review findings contain issues addressing a mode. Defect types in which modes played a major role were completeness and unambiguity. Based on these results, we argue that explicitly specifying modes prior to requirements formulation may increase the artifact quality of natural language requirements specifications.
CITATION STYLE
Vogelsang, A., Femmer, H., & Winkler, C. (2016). Take care of your modes! an investigation of defects in automotive requirements. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9619, pp. 161–167). Springer Verlag. https://doi.org/10.1007/978-3-319-30282-9_11
Mendeley helps you to discover research relevant for your work.