A Conceptual Introduction to Hetero-functional Graph Theory for Systems-of-Systems

TL;DR

This paper introduces hetero-functional graph theory (HFGT) as a conceptual framework bridging systems engineering and network science, enabling interdisciplinary analysis of complex systems through ontological and graph-based models.

Contribution

HFGT provides a novel, linguistically grounded meta-architecture that integrates MBSE and network analysis, facilitating interdisciplinary modeling and analysis of complex systems.

Findings

HFGT preserves heterogeneity of system concepts and functions.

It offers multiple graph-based data structures for analysis.

The modeling process is grounded in ontological and linguistic foundations.

Abstract

A defining feature of twenty first century engineering challenges is their inherent complexity, demanding the convergence of knowledge across diverse disciplines. Establishing consistent methodological foundations for engineering systems remains a challenge -- one that both systems engineering and network science have sought to address. Model-based systems engineering (MBSE) has recently emerged as a practical, interdisciplinary approach for developing complex systems from concept through implementation. In contrast, network science focuses on the quantitative analysis of networks present within engineering systems. This paper introduces hetero-functional graph theory (HFGT) as a conceptual bridge between these two fields, serving as a tutorial for both communities. For systems engineers, HFGT preserves the heterogeneity of conceptual and ontological constructs in MBSE, including system form, function, and concept. For network scientists, it provides multiple graph-based data structures enabling matrix-based quantitative analysis. The modeling process begins with ontological foundations, defining an engineering system as an abstraction and representing it with a model. Model fidelity is assessed using four linguistic properties: soundness, completeness, lucidity, and laconicity. A meta-architecture is introduced to manage the convergence challenges between domain-specific reference architectures and case-specific instantiations. Unlike other meta-architectures, HFGT is rooted in linguistic structures, modeling resources as subjects, system processes as predicates, and operands-such as matter, energy, organisms, information, and money-as objects. These elements are integrated within a system meta-architecture expressed in the Systems Modeling Language (SysML). The paper concludes by offering guidance for further reading.