Automating resilience: how AI network management is redefining connectivity

Improving critical digital infrastructure on which we all depend

Stable, high-quality telecommunications networks are vital infrastructure upon which the global economy depends. Seemingly small increases in their capability and scale can have a significant impact. For example, according to the International Telecommunication Union, a 10% increase in fixed-line broadband penetration can boost GDP by 0.77%, while the same increase in mobile broadband penetration can raise GDP by up to 1.50%. Outages and service disruptions are therefore costly for all stakeholders - yet networks are still vulnerable to such failures, leading to lost productivity and consumer dissatisfaction.

CSPs have historically relied on manual network management techniques—an approach increasingly inadequate for the scale and complexity of modern demands. The industry is now on the cusp of transformation, as CSPs seek to reduce operational costs, enhance resilience, and meet the demands of real-time connectivity. AI network management solutions promise to address these challenges, automating troubleshooting and minimizing downtime by up to 80%, while reducing operational costs by 50%. These technologies also enable delivery of connectivity-as-a-service (CaaS), fostering new business opportunities and improving customer loyalty in an era when user expectations are rising rapidly.

How AI network management works

The AI-driven network automation for traffic and service resilience Catalyst tackles these challenges with solutions rooted in Levels 4 and 5 network autonomy. The approach combines an AI-powered network automation layer with the integration of TM Forum's Open Digital Architecture (ODA) and an array of Open APIs. TM Forum assets are crucial, offering standardized interfaces that ensure seamless integration across diverse network components.

When combined with AI, these elements enable intelligent, intent-based network management spanning from the cloud to the edge. The project team has successfully employed them alongside generative AI (genAI) to enable zero-touch operations, interpreting user intents and translating them into actionable network configurations. For instance, the Service Catalog Management API (TMF620) facilitates real-time service provisioning, adapting dynamically to evolving network conditions and user demands. Similarly, the Resource Inventory API (TMF639) allows the AI layer to automate resource allocation, ensuring rapid adaptation to changing requirements and minimizing the need for manual intervention.

The solution also excels in intent-based management. The Intent Management API (TMF921) empowers CSPs to interpret user or business requirements—such as demands for low latency or increased bandwidth—and configure the network to meet these needs in real time. Built-in conflict resolution ensures that, when multiple requests compete for limited resources, the system prioritizes tasks based on predefined rules or operational parameters.

In cases of anomalies, the Alarm Management API (TMF814) detects issues and initiates automated corrective actions, enhancing network resilience and minimizing downtime. These Open APIs enable the system to achieve a modular and scalable architecture, promoting flexibility while accelerating the deployment of new services.

The value of AI network management

The solution delivers numerous demonstrable benefits, including a substantial reduction in manual interventions, improved service reliability, and enhanced efficiency in resource utilization. Its intent-driven framework resolves conflicts by prioritizing network activities based on predefined business objectives and user requirements.

As Telin’s Delfi Gunardy and Burhan Adi Wicaksana agree with Takayuki Nakamura from NTT, “by establishing a benchmark for autonomous operations, this innovation, when widely deployed, has the potential to create a truly next-generation foundation for commercial solutions that depend on technologies such as 5G, IoT, and edge computing. These advancements require highly responsive networks capable of adapting in real time—a demand this AI network management solution addresses directly.”

Strategically, these innovations support CSPs in their ambition to become providers of advanced, customizable CaaS solutions. This new role within the B2B2X ecosystem enables enterprises to tailor connectivity to their specific needs, unlocking opportunities in sectors such as healthcare, logistics, and smart cities.

Finally, the project is aligned with broader social and environmental objectives, such as the UN SDGs. By improving energy efficiency, extending equipment lifespans, and reducing waste, the solution significantly minimizes the carbon footprint of network operations. In an era of heightened environmental accountability, next-generation solutions like this must prioritize sustainability not only as a social responsibility but also as a critical factor for commercial success.