Digital requirements engineering with an INCOSE-derived SysML meta-model

TL;DR

The paper addresses the mismatch between traditional requirements engineering and SysML-defined system architectures, proposing an INCOSE-derived MBSR extension that embeds GtWR and ISO29148 pattern-based requirement statements into SysML. It implements a SysML profile with a comprehensive meta-model (Attributes, Rules, Characteristics) and a metric suite to support requirements V&V, validating the approach with a NASA JPL Mars Sample Return case study. Results indicate improved requirement expression quality and traceability, while also highlighting automation and tooling challenges that limit cycle times. By delivering open-source profiles and real-world demonstrations, the work advances Digital Engineering goals through deeper integration of RE with MBSE and a centralized authoritative data store (ASoT).

Abstract

Traditional requirements engineering tools do not readily access the SysML-defined system architecture model, often resulting in ad-hoc duplication of model elements that lacks the connectivity and expressive detail possible in a SysML-defined model. Further integration of requirements engineering activities with MBSE contributes to the Authoritative Source of Truth while facilitating deep access to system architecture model elements for V&V activities. We explore the application of MBSE to requirements engineering by extending the Model-Based Structured Requirement SysML Profile to comply with the INCOSE Guide to Writing Requirements while conforming to the ISO/IEC/IEEE 29148 standard requirement statement patterns. Rules, Characteristics, and Attributes were defined in SysML according to the Guide to facilitate requirements definition, verification & validation. The resulting SysML Profile was applied in two system architecture models at NASA Jet Propulsion Laboratory, allowing us to assess its applicability and value in real-world project environments. Initial results indicate that INCOSE-derived Model-Based Structured Requirements may rapidly improve requirement expression quality while complementing the NASA Systems Engineering Handbook checklist and guidance, but typical requirement management activities still have challenges related to automation and support in the system architecture modeling software.

Figures (11)

• Figure 1: Classical SysML requirements modeling using standard relationships to system model elements.
• Figure 2: SysML Stereotype definitions for Requirement Expression and Requirement Set. Refer to Fig. \ref{['fig:GtWR-Attributes']} for further detail on the INCOSE GtWR Attributes.
• Figure 3: Relation Map of example MBSR and related system architecture SysML elements.
• Figure 4: SysML meta-model of INCOSE-derived MBSR classifiers derived from Figure 4 of GtWR2023.
• Figure 5: Model-based INCOSE GtWR Attributes of Well-Formed Needs and Requirements. Refer to Fig. \ref{['fig:Stereotypes']} for further relationship definitions, and Fig. \ref{['fig:Customizations']} for customizations.