Slides: IPEC2021: The Webel Parsing Analysis recipe for text-driven Model-Based Systems Engineering (MBSE) with Systems Modeling Language (SysML)

The Webel Parsing Analysis recipe for SysML is a technique for traceable elicitation of SysML model elements from text extracts quoted from domain source documents. Each text extract is encapsulated in a Snippet extension of the SysMLv1.x ElementGroup stereotype. A Snippet must have a unique domain source Document with a known URI.

The recipe has been applied to substantial real-world applications such as modelling of radio telescopes, particle accelerators, neutron beam instruments, electronics, robotics, astronomical instrumentation, green rating schemes for buildings, and for digital twin encapsulation of building and construction projects.

The Webel Parsing Analysis recipe is a meta-process that can be combined with other model-based systems engineering methodologies and help amplify other approaches through Bridging Diagrams that leverage Parsing Analysis diagrams.

The recipe is particularly useful in combination with requirements engineering; a modelling layer of elicited model elements that includes elicited Requirements is created before direct treatment of the Requirements using the SysML approach. Text statements of requirements can also be used in Snippets for elicitation of general SysML model elements, including those that must Satisfy a Requirement.

The recipe has benefits for many aspects of the modelling process; because the modelling is driven by small text snippets, Parsing Analysis Diagrams have a natural scope. Elicited model elements are then integrated progressively step-wise into robust SysML models that are guaranteed to converge as long as the chosen domain source documents are coherent. The technique also offers a highly effective way of identifying and tracking inconsistencies in source domain documents.

The author suggests that the technique offers a powerful cognitive advantage over purely graphically based modelling that is not related to source texts; the human brain understands more when text is presented alongside graphical symbols, which approach also helps make SysML models comprehensible to more stakeholders. The modelling recipe as for SysMLv1.6 and as tuned for the MagicDraw SysML and Cameo Systems Modeler tools is presented, with a word on adapting the recipe for future SysMLv2.