An Empirical Quality Assessment of Automotive Use Cases

As functionality in vehicles grows more complex and development becomes distributed over several geographical sites, elicitation and visualization of requirements become more critical. This paper presents a set of evaluation criteria for the quality of use cases. The criteria are applied to use cases that are currently used in industry and developed according to current industrial practice. The paper presents statistics of quality defects that occur in industry and proposes academic solutions that may be applied to solve them. The study is based on 43 use cases from Volvo Car Corporation spanning three different function areas of a vehicle. The most common quality defect classes of the evaluated use cases are missing elements, irrelevant steps, incorrect linguistics and level of detail. Furthermore, it is concluded that a common taxonomy and cross team reviews are needed to further improve the quality and usefulness of use cases