A Model-driven and Template-based Approach for Requirements Specification

Requirements specification and verification play an important role in the certification of safety-critical software (SCS). These activities are costly and error-prone because SCS exhibit a high number of requirements and most SCS manufacturers are still using natural language to specify these requirements. On one hand, natural language can introduce ambiguity and inconsistency. On the other hand, formal languages add an overhead to the requirements specification because of their complexity. Controlled Natural Languages (CNLs) fill these gaps by offering a middle-ground solution, although not yet well adopted by the industry. In this paper, we introduce an approach that combines CNLs and model-driven engineering (MDE) for requirements specification. The approach was proposed to support an industrial partner in the certification process of a SCS. Our approach uses templates and relies on two types of models: models that specify the templates, and a model of the domain of the system at hand. Using models of the templates enables to automate some requirements analysis tasks. Using a domain model allows the auto-completion and verification of requirements specified using the templates. We implemented the approach and validated it using three case studies and more than a thousand requirements. We observed that our approach and underlying templates are applicable across domains and that the templates yield requirements with better quality in terms of necessity, ambiguity, completeness, singularity, and verifiability.