Safety-critical systems are typically described using high-level models, as they provide a more precise representation compared to traditional document-centric approaches. However, high-level engineering models are only beneficial if their structure and the correctness of the behavior they describe can be trusted. Since manually examining and ensuring this is costly, system engineers are increasingly supported by automated verification and validation (V&V) tools.

Although such tools exist, their development and implementation are extremely costly because supporting new languages often requires re-implementing everything from scratch. This is due to the minor nuances of different languages that hinder reusability.

To address this issue, a novel approach involves the application of semantic models, which store the structural and behavioral characteristics of a language in an external, easily replaceable and modifiable model, effectively providing "scripting" capabilities over the V&V tool. The advantage of this approach is that supporting new languages or modifying existing tools can be achieved without altering or recompiling the source code of the specific V&V tool, significantly reducing the required resources.

In my research, I explore the applicability of the semantic modeling approach using the Semantifyr framework for both classical and ontological languages, with a special focus on: 1) the efficiency of the V&V process in determining appropriate behavioral abstractions; and 2) supporting new and existing modeling languages with V&V tools.