New Ontology Enhances Distributed AI Workflows Across Edge-Cloud

Viorica Rozina Chifu, Tudor Cioara, Vasile Ofrim, Liana Toderean, Ionut Anghel, Laura Daniele, Cornelis Bouter· August 28, 2026 View original

Key takeaways

  • Distributed AI workflows across edge-fog-cloud environments face interoperability and orchestration challenges.
  • A new SAREF-compliant ontology provides a unified semantic model for AI pipelines and infrastructure.
  • The ontology enables automated reasoning and resource-aware orchestration of distributed AI applications.
  • It significantly improves deployment success rates and reduces orchestration decision times.

Who benefits

IoTSmart CitiesManufacturingEnergyTelecommunications

Summary

A new SAREF-compliant ontology is proposed to improve interoperability and orchestration of distributed AI workflows across heterogeneous edge, fog, and cloud environments. It extends SAREF4SYST to model AI pipelines, resources, and deployment constraints, enabling unified semantic representation and automated reasoning for AI applications.

Current semantic models struggle to adequately represent distributed AI workflows and their execution across diverse computing environments, including edge, fog, and cloud infrastructures. This often leads to incompatible descriptions of AI processes and resources, hindering interoperability, orchestration, and reuse of AI components. To overcome these challenges, a new SAREF-compliant ontology has been developed. This ontology extends the existing SAREF4SYST framework by incorporating concepts for modeling AI pipelines, executable AI jobs, computational resources, deployment constraints, and communication relationships. This creates a unified semantic model that encompasses both AI workflows and the heterogeneous computing infrastructure. The proposed ontology facilitates semantic interoperability, automated reasoning, and resource-aware orchestration of distributed AI applications, all while aligning with the ETSI SAREF ecosystem. Evaluated through smart grid energy services orchestration scenarios, the ontology demonstrated high deployment success rates (90-100%) and rapid orchestration decision times (below 80 ms), proving its effectiveness in managing complex AI deployments across various environments.

Why it matters

Professionals in AI engineering, IoT, and cloud architecture can leverage this ontology to design and manage more interoperable, efficient, and robust distributed AI systems across diverse computing landscapes.

How to implement this in your domain

  1. 1Explore the SAREF-compliant ontology for modeling distributed AI workflows in edge-fog-cloud environments.
  2. 2Adopt semantic modeling principles to describe AI pipelines, resources, and deployment constraints for improved interoperability.
  3. 3Integrate automated reasoning capabilities into AI orchestration platforms using the proposed ontology.
  4. 4Pilot the ontology in a specific distributed AI project to evaluate its impact on deployment success and orchestration efficiency.

Original post by Viorica Rozina Chifu, Tudor Cioara, Vasile Ofrim, Liana Toderean, Ionut Anghel, Laura Daniele, Cornelis Bouter

"arXiv:2608.26160v1 Announce Type: new Abstract: Nowadays semantic models provide limited support for representing distributed AI workflows and their execution across heterogeneous edge, fog, and cloud environments. Therefore, AI processes and resources are often described using i…"

View on X

Originally posted by Viorica Rozina Chifu, Tudor Cioara, Vasile Ofrim, Liana Toderean, Ionut Anghel, Laura Daniele, Cornelis Bouter on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses

More in AI Engineering & DevTools