Large-scale control systems, as encountered in the domain of road traffic management, typically deal with highly-dynamic environments providing information about a large number of real-world objects, which stem from multiple heterogeneous sources and are anchored in time and space. Human operators of such systems face information overload which endangers the recognition of critical situations. Situation awareness systems should support operators fulfilling their tasks by leveraging their awareness of the ongoing situations. However, current approaches to SAW miss a common conceptual model necessary for various aspects of SAW. Although the application of ontologies for filling this gap has been proposed in recent years, ontology-driven SAW systems are nevertheless still in their infancy. In this paper, we shape the vision of an ontology-driven SAW system by the analysis of application scenarios facilitating the features of formal ontologies. We illustrate the suggested scenarios with examples from the field of road traffic management and argue that an ontology-driven SAW system does not replace but may actually enhance traditional probabilistic approaches to SAW.