XAI Verifies ML Decisions in Optical Networks

Omran Ayoub, Carlos Natalino, Ali Al Housseini, Felix Foschum, Philipp Morger, Tiziano Leidi, David Hock, Paolo Monti· July 24, 2026 View original

Summary

This work introduces explanation-based runtime verification, a method that uses XAI techniques to assess the soundness of individual machine learning decisions in optical network control loops before execution. It evaluates explanation coherence and physics grounding consistency to intercept erroneous decisions and enhance trustworthiness.

Machine learning models are increasingly used in optical network automation for tasks like failure management and resource allocation. However, incorrect ML decisions can severely impact service quality and network stability, making reliability at deployment time crucial. To address this, researchers propose explanation-based runtime verification. This approach leverages Explainable AI (XAI) techniques to scrutinize individual ML predictions before they trigger control-plane actions. It assesses the coherence of explanations and their consistency with physical network principles. By doing so, the system can identify and defer or reject uncertain decisions, significantly improving the trustworthiness of ML-driven automation. Demonstrated on lightpath quality of transmission classification, the method effectively intercepts a substantial portion of erroneous decisions while maintaining high automation rates.

Why it matters

This innovation provides a critical safety net for AI deployments in sensitive infrastructure like optical networks, ensuring that automated decisions are reliable and trustworthy, thereby preventing costly service disruptions.

How to implement this in your domain

  1. 1Integrate XAI-based runtime verification into ML-driven automation pipelines for critical infrastructure.
  2. 2Develop mechanisms to assess explanation coherence and physics grounding for AI decisions.
  3. 3Establish protocols for deferring or rejecting ML decisions flagged as uncertain by the verification system.
  4. 4Pilot this approach in a controlled environment to validate its effectiveness in intercepting errors.

Who benefits

TelecommunicationsUtilitiesCritical InfrastructureManufacturingDefense

Key takeaways

  • Explanation-based runtime verification enhances trustworthiness of ML in optical networks.
  • XAI techniques are used to assess the soundness of individual ML decisions before execution.
  • The system evaluates explanation coherence and physics grounding consistency.
  • It can intercept erroneous decisions while maintaining high automation rates.

Original post by Omran Ayoub, Carlos Natalino, Ali Al Housseini, Felix Foschum, Philipp Morger, Tiziano Leidi, David Hock, Paolo Monti

"arXiv:2607.20675v1 Announce Type: new Abstract: Machine learning (ML) models are increasingly integrated into optical network automation frameworks to support tasks such as failure management, performance monitoring and resource allocation. In these environments, ML-driven predic…"

View on X

Originally posted by Omran Ayoub, Carlos Natalino, Ali Al Housseini, Felix Foschum, Philipp Morger, Tiziano Leidi, David Hock, Paolo Monti on X · view source

Want to go deeper?

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

Explore courses