AI-Driven Optical Transport Networks: Architectures, Applications, and Intelligent Automation

SC463 - AI-Driven Optical Transport Networks: Architectures, Applications, and Intelligent Automation

16 Mar 2026
08:30 - 12:30

Short Course Level

Intermediate

Short Course Description

This short course covers practical applications of T-SDN and AI-enabled automation in optical transport networks, extending classical SDN concepts with ML, LLMs, and agentic AI. Using a problem-solution approach, we explore how service providers, hyperscalers, and enterprises can leverage these technologies to address operational challenges such as rapid service turn-up, fault detection/analysis,
and resource optimization.

Part 1 introduces the T-SDN architecture, data models, protocols, and frameworks. We examine how open interfaces and streaming telemetry enable software-based control and intent-driven automation. The session maps standards, industry initiatives and open source projects to the overall architecture and shows how AI/ML pipelines (feature engineering, MLOps/AIOps, guardrails) enhance model-based approaches.

Part 2 explores commercial use cases from communication service providers and data center operators. Starting with the optical layer, we examine how T-SDN and AI support open line systems and disaggregated WDM platforms. We then address multilayer scenarios (packet-optical and IP-over- WDM), including cross-layer path computation, multi-layer resilience, and multi-domain interoperability.

Throughout the course, we focus on operations-ready AI:
• Predictive maintenance and soft failure detection
• ML-based health monitoring, anomaly localization, and capacity headroom determination
• LLM-assisted intent translation and what-if scenario planning
• Policy-controlled agentic AI and ChatOps for human-in-the-loop workflows
Key words: SDN, transport, optical networks, open platforms, multi-layer networks, multi-domain
networks, packet-optical, AI/ML, LLMs, network automation, network intelligence
• Adtran Confidential

Short Course Benefits

By the end of this course, you will be able to:

• Understand key T-SDN concepts, data models, protocols and open‐source frameworks (OpenROADM, ONF T-API, OpenConfig/YANG, gNMI/NETCONF)
• Discuss how AI/ML augments T-SDN to address key operator challenges
• Summarize how telemetry and AI pipelines integrate with network controllers and operational/business support systems (OSS/BSS)
• Understand the role of ChatOps and LLM agents to simplify service provisioning, troubleshooting, and network change management
• Assess operational risks and benefits of agentic AI, including explainability, safety, governance, and compliance logging for AI-generated actions
• Orchestrate cross-layer (packet-optical, IP-over-WDM) and multi-domain automation using open interfaces, intent frameworks, and policy constraints.
• Apply predictive maintenance and soft-failure detection using optical performance data to drive proactive operations and reduce mean time to repair (MTTR).

Short Course Audience

This course is designed for network and system architects, engineers, researchers, and technical managers working on packet-optical network automation, multilayer optimization, and data-driven operations. Attendees should have a basic understanding of SDN concepts and be interested in deepening their knowledge of architectures, applications, and implementation.

• Achim Autenrieth

  ADVA Optical Networking SE, Germany

  About the Instructor

  Achim Autenrieth is Senior Director Advanced Technology at Adtran Networks SE, where he is leading the research activities on networking technologies including intelligent network automation and SDN control of 5G/6G and disaggregated optical transport networks, machine learning, and planning and evaluation of multilayer optical networks. Achim is member of IEEE and VDE/ITG, he authored or co- authored more than 120 reviewed and invited scientific publications and he is technical program committee member of OFC (2018-2021, 2024-2026), ECOC (2011-2017), ONDM, DRCN and RNDM.

• Jörg-Peter Elbers

  ADVA Optical Networking SE, Germany

  About the Instructor

  Jörg-Peter Elbers is VP of Advanced Technology, Standards & IPR at Atran, responsible for global technology strategy, applied research, standardization, and intellectual property. He has over 25 years of experience in the telecommunications industry, with previous technical leadership roles at ADVA, Ericsson, Marconi, and Siemens. Jörg holds a Dr.-Ing. (PhD) degree in Electrical Engineering from the Technical University of Dortmund.