RuntimeVerifikation

RuntimeVerifikation, in English usually called runtime verification, is a field of computer science that focuses on observing the behavior of a running system to ensure it satisfies formal specifications formulated before execution. Unlike exhaustive verification, which analyzes all possible executions, runtime verification checks actual executions as they occur and can report violations or incomplete results.

Monitors in RuntimeVerifikation observe events such as method calls, messages, and state changes, and compare them against temporal specifications. Specifications are typically written in temporal logics (for example linear temporal logic or metric temporal logic) or in automata-based formalisms, regular expressions over event streams, or pattern-based languages. Monitors may operate online (during execution) or offline (postmortem analysis) and can take into account real-time constraints.

RuntimeVerifikation complements model checking and static analysis by focusing on runtime evidence rather than exhaustive proof. It is used for detection, diagnosis, and sometimes enforcement through monitors that can trigger corrective actions. In practice, there is often a trade-off between completeness and practicality, as monitoring must be efficient and incur only modest overhead on the running system.

Applications of RuntimeVerifikation span safety-critical and embedded domains such as aerospace, automotive, medical devices, distributed cloud services, and consumer electronics. Properties of interest include safety, liveness, and timing requirements, as well as reliability and fault detection. The approach is used throughout development, testing, and production to detect property violations, aid debugging, and support runtime assurance.