The world of complex engineering systems faces a persistent challenge: the gap between meticulous design intentions and the unpredictable realities of operational use. A new framework aims to bridge this divide, offering a structured way to weave operational insights directly into the design process for safer, more reliable outcomes.
Published in the journal Computers, the paper “Integrated Systems Ontology (ISOnto): Integrating Engineering Design and Operational Feedback for Dependable Systems” introduces a sophisticated semantic model designed to enhance dependability across the system lifecycle. The work brings together expertise from multiple institutions and industries, presenting a practical approach to knowledge integration that could reshape how engineers approach failure prevention and system resilience.
Understanding the Core Challenge in Dependable Systems
Dependable systems engineering focuses on creating products and infrastructures that perform reliably under expected and unexpected conditions. Industries such as automotive, aerospace, rail, and energy rely on this discipline to minimise downtime, protect users, and meet stringent safety standards.
Traditional approaches often treat design and operation as separate phases. Engineers create detailed models during development, then hand over systems to operators who collect real-world performance data. Feedback rarely flows back efficiently, leading to repeated mistakes, incomplete failure mode analyses, and slower improvements in subsequent generations of products.
This disconnect becomes costly when safety-critical components fail in the field. Missed opportunities to learn from operational data mean that known vulnerabilities persist across product iterations, while emerging issues surface only after costly incidents.
What Makes ISOnto Different
ISOnto represents a formal ontological framework specifically engineered to link design artefacts with operational experience. An ontology in this context is a structured representation of knowledge that defines key concepts, their relationships, and rules for reasoning about them.
By creating a shared semantic language, ISOnto allows design models—such as function-behaviour-structure representations—and operational records—like failure reports, maintenance logs, and sensor data—to coexist within a single, machine-readable structure.
The framework supports both validation of existing failure modes and discovery of previously unrecognised ones. It draws on established concepts from Model-Based Systems Engineering while extending them with explicit mechanisms for continuous learning from field data.
The Authors and Their Expertise
The research is led by Haytham Younus, a researcher at Cranfield University with a focus on AI-enabled reliability frameworks. Co-authors include Felician Campean from the University of Bradford, Sohag Kabir from the University of Huddersfield, and industry experts Pascal Bonnaud and David Delaux from Renault. This blend of academic rigour and industrial insight ensures the ontology addresses both theoretical soundness and practical applicability.
The collaboration reflects a growing trend of university-industry partnerships aimed at solving real-world engineering problems through advanced knowledge representation techniques.
Key Components of the ISOnto Framework
ISOnto builds upon earlier work by the team on function-behaviour-structure-failure mode ontologies. It organises knowledge into layered representations that capture:
- Core system functions and their intended behaviours
- Potential failure modes and their effects at multiple levels of decomposition
- Operational contexts, usage profiles, and environmental factors
- Real-world performance data and observed anomalies
The ontology uses semantic technologies to enable automated reasoning. When new operational data arrives, the system can check consistency with design assumptions, flag discrepancies, and suggest updates to failure mode analyses.
Photo by Markus Winkler on Unsplash
Practical Application: The Headlamp Case Study
The paper demonstrates ISOnto through a detailed case study involving an automotive headlamp system. Engineers decomposed the system into functional and structural elements, mapped known failure modes, and then integrated anonymised field failure data.
Results showed improved traceability between design choices and observed issues. The framework helped identify patterns that traditional spreadsheet-based FMEA processes often overlook, such as recurring failures linked to specific usage environments or maintenance practices.
This example highlights how ISOnto supports more dynamic and data-driven reliability engineering workflows.
Benefits for Industry and Research
Organisations adopting ISOnto could see several advantages:
- Faster identification of design improvements based on actual usage
- More comprehensive and living FMEA documents that evolve with operational experience
- Better support for regulatory compliance through traceable safety arguments
- Foundation for AI tools that automate parts of reliability analysis
In research settings, the ontology provides a common language that facilitates collaboration across disciplines and institutions. It also supports comparative studies between different systems and industries.
Connection to Broader Trends in Systems Engineering
ISOnto aligns with the shift toward Model-Based Systems Engineering and digital twins. By formalising the link between design models and operational feedback, it contributes to the vision of continuously learning, self-improving engineering processes.
The framework also complements emerging standards in safety and reliability that emphasise lifecycle thinking and knowledge management.
Challenges and Considerations for Adoption
Implementing an ontology-based approach requires investment in data quality, tool integration, and organisational change. Companies need skilled personnel who understand both systems engineering and semantic technologies.
Data privacy and intellectual property concerns may arise when sharing operational information across supply chains. The framework’s modular design aims to address some of these issues by allowing selective sharing of knowledge elements.
Future Directions and Research Opportunities
The authors outline several avenues for extending ISOnto. These include tighter integration with digital twin platforms, expansion to cover entire product families or fleets, and enhanced reasoning capabilities through machine learning.
Further case studies in aerospace, rail, and energy sectors would strengthen validation. Collaboration with standards bodies could help embed the ontology in industry toolchains.
Implications for the Wider Engineering Community
Beyond immediate applications, ISOnto contributes to the growing body of work on knowledge representation in engineering. It demonstrates how ontologies can move from academic concepts to practical tools that address longstanding industry pain points.
Early adopters in automotive and related sectors may gain competitive advantages through improved product reliability and reduced warranty costs. The open nature of the publication encourages broader experimentation and refinement by the research community.
Conclusion
The introduction of Integrated Systems Ontology (ISOnto) marks a significant step toward closing the design-operation feedback loop in dependable systems engineering. By providing a semantic bridge between models and real-world performance, the framework offers a foundation for more resilient, learnable, and ultimately safer engineering outcomes.
As industries continue to embrace digitalisation and data-driven methods, approaches like ISOnto will become increasingly valuable. Researchers and practitioners interested in systems dependability now have a concrete starting point for exploring semantic integration in their own work.
The full paper is available as open access, allowing the global engineering community to examine the details and build upon this promising foundation.
