A Method and Tool for Automated Analysis of Heavy Vehicle RequirementsA Method and Tool for Automated Analysis of Heavy Vehicle Requirements

Abstract: The introduction of the functional safety standard ISO 26262 was motivated by an increasing demand to ensure reliability and correctness of safety-critical systems in the automotive industry. However, the adoption of this standard in the automotive industry is hindered by a number of obstacles. Scania is an industrial partner in the VeriSpec project which studies these obstacles and proposes relevant tools and methods compliant with academic and industrial needs. This thesis is within the scope of the VeriSpec project, and aims to address one of the project’s goals, which is to provide tool support for a pattern-based requirement formalization process. The Specification Property System (SPS) proposed by Konrad and Cheng is a patterning method that provides automatic translation of system properties into temporal logics. The SPS also helps in restricting the introduction of ambiguities and inconsistencies in system specification properties. However, the adoption of the SPS in the industry is hindered due to some issues. These issues are, a long learning curve, Constrained Natural Language (CNL) ambiguities, and the lack of tool-support for real-time SPS patterns. In this thesis, a qualitative research study with a literature survey has been performed to find and select state-of-the-art supportive methods to provide feedback on the formalized requirements’ semantics. The Scania Specifier tool has been extended and modified to support a requirement formalization process using the SPS qualitative and real-time patterns. In addition, three supportive methods that resulted from the research study have been integrated into the Specifier tool to provide different feedback options for the users. Finally, the performance of the Specifier tool and the feedback of the supportive methods have been evaluated. The outcome of the study shows that the feedback of the supportive methods helped in guaranteeing the intended behavior of the requirement developers. In addition supportive methods’ feedback enhanced user-friendliness, and aided the users in shortening the SPS learning curve. Finally, an additional outcome of the study is in the form of a number of suggestions and emerged patterns with regard to the SPS usage and supportive methods’ feedback.