Improving Development of Communication Software in Industrial Control Systems using Simulation

In the industrial domain customers expect a product longevity of 10-20 years, with high reliability and availability. Since industrial distributed control systems often are safety critical, aspects such as determinism, low latency and jitter are crucial. More and more industrial systems are becoming connected to the Internet, since end customers are requiring e.g. business intelligence and diagnostic information, anywhere at any time. Industrial systems that traditionally have been isolated are now facing entirely new challenges that will require new competences and ways of working. Introducing a new type of network in the industrial domain is a big investment, with high risks, often lacking known best practices. Time to market with sufficient quality is of high importance. A lot of time is spent on isolated activates, such as, simulations, updating tools, collecting requirements, design, coding, debugging, documentation, creating testbeds, validation and reviews. Therefore, there is a need to improve the efficiency when moving between the research and development phases for several reasons, e.g., integrate innovative research findings into industrial systems, shorten time to market, and improve product quality. This thesis focuses on improving efficiency during research and development of communication software. First, network evaluation methods are studied, and key industrial challenges are identified. For example, despite a huge research effort on network simulators and virtualization, there are still challenges that need to be addressed, in order for increased industrial benefits. Secondly, this thesis propose a flexible communication stack design that supports different run-time behaviors, from real-time operating system to bare-metal systems without an operating system, and different types of communication protocols, from real-time to non-real-time. Finally, this thesis propose a set of key features from network simulators, that are implemented and used as a case study in a research project. These contributions lead to simplification and increased automation, hence reducing the amount of manual work during research and development.