| |
Monitoring and control solutions for large one-off systems are typically built in silos using multiple tools and technologies. Functionality such as data processing logic, alarm handling, UIs, device drivers are implemented by manually writing configuration code in isolation and their cross dependencies maintained manually. The correctness of the created specification is checked using manually written test cases. Non-functional requirements – such as reliability, performance, availability, reusability and so on – are addressed in ad hoc manner. This hinders evolution of systems with long lifetimes. For ITER, we developed an integrated specifications environment and a set of tools to generate configurations for target execution platforms, along with required glue to realize the entire M&C solution. The SKA is an opportunity to enhance this framework further to include checking for functional and engineering properties of the solution based on domain best practices. The framework includes three levels: domain-specific, problem-specific and target technology-specific. We discuss how this approach can address three major facets of complexity: scale, diversity and evolution.
|
|