Skip to Main Content
Decision makers in safety-critical domains rely on data from dependable information systems. Traditional time- and safety-critical information systems, such as traffic management systems, have been using proprietary point-to-point data links, which are very dependable, but also time-consuming and costly to change due to the need to manually adapt a multitude of deployed systems. In this paper we introduce a model-driven architecture (MDA) system approach that describes explicitly stakeholder quality requirements on dependable data links between systems for decision support and generates new system versions that implement these requirements. The MDA approach is expected to a) improve the quality (assurance) of system requirements; b) support more explicit feedback on the quality of intermediate models during systems development; and c) provide better auditing capabilities of the systems development process. Based on an industry case study we describe the MDA concept of the system, the development process, and how software quality can be measured and improved.