Non-technological barriers: the last frontier towards AI-powered intelligent optical networks

Area of Science:

• Telecommunications Engineering
• Computer Science
• Network Science

Background:

• Machine learning (ML) has achieved significant success across various scientific and technological domains.
• ML is considered a transformative technology for modernizing optical networks into intelligent, autonomous systems.
• Despite extensive research, ML has not achieved widespread adoption in commercial optical networks.

Purpose of the Study:

• To identify and analyze the critical non-technological barriers hindering ML adoption in optical networks.
• To emphasize the need for addressing these overlooked factors for the successful implementation of ML in optical networks.

Main Methods:

• Perspective-based analysis of current ML applications in optical networks.
• Identification of key non-technological challenges impacting real-world deployment.
• Discussion of stakeholder responsibilities in resolving these challenges.

Main Results:

• The primary obstacle to ML adoption in optical networks is the lack of focus on non-technological issues.
• Critical factors such as data management, standardization, and ethical considerations remain unaddressed.
• These non-technological gaps prevent the realization of ML-powered autonomous optical networks.

Conclusions:

• Widespread adoption of ML in commercial optical networks requires a paradigm shift to address non-technological challenges.
• Resolving issues related to development and deployment is essential for creating intelligent and autonomous fiber-optic systems.
• Collaboration among stakeholders is necessary to overcome these barriers and achieve the vision of next-generation optical networks.