Dominic Schneider explores transformer-based prognostics for optical amplifier reliability at OFC 2026
Session will examine edge-level AI health prediction for optical amplifiers and how it supports uptime, SLA assurance and lower operating cost
- Speakers: Dominic Schneider, senior research engineer of advanced technology at Adtran
- Topics: Poster Th2A.23: Transformer-based prognostics: enhancing network availability by improved monitoring of optical fiber amplifiers
- When: Thursday, March 19, 10:30 a.m. PT
- Where: OFC 2026, Los Angeles Convention Center, Los Angeles, California, USA https://www.ofcconference.org
Optical networks must deliver high availability as complexity and cost pressure increase. Reactive maintenance of optical fiber amplifiers can drive avoidable outages, SLA risk and inefficient field activity. Operators need interoperable, automation-ready monitoring that turns condition data into reliable insight. Predictive approaches help reduce operational risk, improve ROI and support more consistent service quality at scale.
In this poster session Dominic Schneider will outline how a lightweight transformer model predicts optical fiber amplifier lifetime using condition-based monitoring data. He will explain how edge-level inference supports real-time maintenance decisions and integrates with operational workflows. Attendees will gain insight into deployable AI that improves time to act, reduces disruption and supports more predictable network availability.