Write Functional Specifications with Models

Writing functional specifications in natural language and then duplicating the effort by creating models introduces unnecessary overhead and risks. This duplication not only increases costs but also creates potential inconsistencies between textual descriptions and their corresponding models.

Moreover, natural language is inherently ambiguous and often ill-suited to describe logical behaviors precisely. For instance, when asked whether electronic cigarettes are prohibited on airplanes, most people say yes—yet in reality, only their use is forbidden, not their presence. Such nuances can be critical.

In this session, CIL4Sys Engineering will demonstrate how elementary functional behaviors can be captured with concise sequence diagrams, and how the aggregation of these behaviors can be structured into state machines. Each modeled behavior can then be linked to a template-based sentence, automatically generating a functional requirement that is consistent with the model.

Participants will also discover a virtual test bench—a functional digital mockup—that enables the simulation, verification, and validation of the modeled behaviors. This approach facilitates better communication, shared understanding, and early validation of system requirements.

The session will follow a hands-on tutorial based on a representative space mission use case. Participants who wish to follow along actively are welcome to do so: information on installing the free version of the tooling and accessing the documented tutorial will be made available prior to the workshop.