Resilient Reactive Systems based on Runtime Semantic Models

IoT, smart cities, cyber-physical systems and sensor networks represent new classes of highly dynamic, complex systems. The behavior of these systems should be designed in order to react to external changes, i.e. they are reactive and context-aware, and also to internal ones to be able to reconfigure themselves for handling possible anomalies. These requirements ask for a runtime representation of application logic and its context, enriched with variation points that associate different behaviors to possible changes.