systems engineering

Conventional requirements engineering tools lack direct access to SysML architecture models, leading to redundant requirement definitions, semantic fragmentation, and broken traceability. Method: This paper proposes an executable, structured requirements metamodel that integrates INCOSE requirements writing practices with SysML modeling capabilities. Strictly aligned with ISO/IEC/IEEE 29148 and INCOSE guidelines, it leverages a SysML Profile extension, an MBSE integration framework, and a compliance rule engine to enable native interoperability between requirements and architecture models. Contribution/Results: The metamodel was deployed and validated on two real-world NASA JPL space systems. It significantly improves requirement semantic completeness and verifiability, enhances coverage of the NASA Systems Engineering Handbook checklist, and—critically—provides the first empirical evidence of rapid improvement in requirements expression quality. The evaluation also identifies key bottlenecks in current toolchains regarding automated support for such integrated practices.

Engineering models (e.g., SysML) lack formal planning semantics—such as preconditions, effects, resource constraints, and temporal bounds—hindering task reachability and performance evaluation across system variants. Method: This paper proposes a model-driven approach natively integrated into SysML, leveraging a custom SysML profile to embed symbolic planning semantics directly into engineering models. It enables fully automated, bidirectional transformation from SysML models to PDDL domain and problem files—without external models or manual intervention—ensuring semantic consistency and model reusability. The method synergistically combines model transformation algorithms with symbolic planning techniques. Contribution/Results: Evaluated on an aircraft assembly case study, the approach validates functional feasibility and execution efficiency across multiple system variants. It significantly enhances interoperability between Model-Based Systems Engineering (MBSE) and AI planning, advancing automation and rigor in early-phase system design and analysis.

This work proposes a lightweight framework to address the high cost and poor contextual retention inherent in traditional engineering decision capture methods. By modeling decision alternatives as slices of system models and embedding them directly into model-based systems engineering (MBSE) workflows, the approach explicitly links decision knowledge with requirements, behavioral elements, and architectural components. This integration significantly reduces the overhead of decision documentation while enhancing the preservation of contextual information. The feasibility of the framework is demonstrated through a case study on aircraft architecture simplification, which confirms its effectiveness in improving decision reusability and integration efficiency within MBSE environments.

This study addresses the common disconnect between SysML models and physical implementations in traditional model-based systems engineering, where models are often abandoned during hardware verification. To bridge this gap, the authors propose a novel bidirectional communication architecture that enables direct, real-time message exchange between executable SysML models and physical hardware—without requiring intermediate translation or co-simulation platforms. By integrating an embedded C++ SysML-side server into IBM Rhapsody and coupling it with a Raspberry Pi hardware interface, the approach embeds SysML statecharts directly into the hardware-in-the-loop verification loop. Validation on a logic gate case study demonstrates perfect output consistency between the SysML model and physical hardware, as confirmed by Karnaugh map comparison. This result substantiates the feasibility of model-driven hardware verification, significantly shortening the digital thread and enhancing model reuse throughout the development lifecycle.

Early in system design, ambiguous information impedes rigorous assessment of model–physical system consistency, completeness, and reusability, thereby compromising quality assurance and cost control. To address this, we propose a SysML extension specifically for cyber-physical systems (CPS), which—uniquely—formalizes SysML as a bidirectionally verifiable language supporting integrity validation. Our method unifies and quantifies simulation-to-physical mappings across structural, behavioral, and temporal dimensions, enabling definable, traceable, and measurable consistency evaluation. It integrates three core techniques: metamodel-driven constraint specification, model difference analysis, and dynamic temporal consistency checking. Evaluated on an industrial CPS case study, the approach achieves over 92% accuracy in automated cross-domain deviation identification and substantially reduces integration verification time.

This work addresses the behavioral gap between formal verification and actual execution in traditional engineering approaches, which often neglect execution semantics. To bridge this semantic divide, the paper proposes a Modeling and Simulation-Based Engineering (MSBE) methodology that explicitly treats execution semantics as a first-class engineering entity. It defines executability as the admissible model space induced by the stabilization of execution conditions and unifies model behavior with physical execution through an iterative cycle of formal execution, experimental execution, verification, and activity-mediated validation. Integrating formal methods, simulation-based verification, activity theory, and constraint modeling, MSBE establishes a general-purpose engineering framework applicable to diverse cyber-physical systems (CPS). The approach demonstrates its generality and effectiveness across four CPS categories: human-centric, biophysical, technological, and digital twin systems.

This study addresses the unclear role of current AI tools in requirements engineering and their alignment with INCOSE standards. Designing a controlled experiment grounded in INCOSE’s “good requirements” criteria, the work systematically compares AI tools and human experts across key quality dimensions—consistency, completeness, clarity, and testability. The findings empirically demonstrate, for the first time, that AI can efficiently and consistently perform initial quality assessments at the syntactic and structural levels, yet still requires expert intervention for higher-order tasks involving contextual understanding, ambiguity resolution, and trade-off reasoning. The results delineate a clear integration pathway wherein AI serves as a decision-support aid rather than a replacement for human judgment, offering both methodological foundations and practical guidance for AI-assisted requirements validation.

This study addresses the limitations of existing SysML verification approaches, which are often tool-dependent and restricted to performance properties, lacking support for automated validation of behavioral and interface requirements. To overcome these shortcomings, this work proposes a tool-agnostic, automated verification workflow driven by SysML test cases, integrating UML Testing Profile and behavioral diagram constructs to enable unified validation of multidimensional attributes—including behavior, timing, and state responses. The methodology was developed through a mixed-methods research strategy combining literature review and stakeholder interviews, and its efficacy was empirically validated across two independent SysML toolchains. The approach not only transcends the constraints of conventional parametric methods but also enables automatic traceability of verification results back to the original model elements.

This work addresses a critical limitation of traditional program repair approaches, which focus solely on code while overlooking the possibility that requirements themselves may be erroneous or outdated, leading to inconsistencies between system implementation and intended specifications. To bridge this gap, the paper introduces the first automated requirement repair framework tailored for Simulink Requirements Tables, shifting the repair target from code to requirements. The framework analyzes system execution traces, handles real-valued temporal signals, evaluates the semantics of declarative requirements, and automatically generates corrective patches. Evaluated across six real-world case studies involving twelve requirements, seven variants of the framework successfully produced correct and meaningful repairs, effectively restoring compliance between requirements and system behavior and addressing a key research gap in the co-evolution of requirements and implementations.

This work addresses the current lack of native integration between model-based systems engineering (MBSE) and data-driven artificial intelligence, which hinders the full-lifecycle development of trustworthy autonomous cyber-physical systems. The paper proposes IDDMBSE, a novel methodology that embeds data-driven feedback loops into each phase of the MBSE V-model through SysML modeling, ROS autonomy stack mapping, and a hybrid architecture, enabling co-design, co-verification, and co-optimization of models and data. IDDMBSE establishes a pioneering systems engineering paradigm that deeply unifies MBSE with AI, supports natively composable domain-specific languages, and delivers an integrated assurance workflow spanning design, evaluation, and runtime verification. The approach is validated on trustworthy ground robots using the open-source toolchain—including PERFECT, TRADES-X, and VERITAS—demonstrating capabilities in sensor selection, risk-aware path planning, formal behavior tree verification, robust perception, multi-robot coordination, and an adversarial terrain testing environment released in Isaac Sim.