The path to autonomous intelligent networks

Telecommunications (telco) service providers face a landscape of massive operational complexity. As they adopt 5G standalone architectures and multivendor radio access networks (RANs), they must manage billions of events across hundreds of thousands of network elements every day. Manual, script-based operations are no longer sufficient to keep pace with these demands.

Transitioning to an autonomous intelligent network is now a foundational requirement for service providers to remain competitive. This shift addresses 3 critical priorities for executive leadership:

• Creating new revenue
• Reducing costs
• Mitigating risk

An open source approach accelerates the integration of new applications and services, provides freedom of choice and access to the latest market innovation, and gives service providers the necessary transparency they need to support regulatory and sovereign compliance.

Charting the journey to full autonomy

The progression toward full autonomy is guided by the TM Forum maturity model which consists of levels from zero to 5. Currently, most service providers operate at level 2—partial automation. The ultimate goal for many is reaching level 4—high automation—where the network detects environmental changes and corrects itself without human intervention.

The TM Forum’s maturity model has the following characteristics and capabilities:

• Decoupled operational layers run business, service, and resource layers independently
• Intent-driven interfaces base communication on what to achieve, not how to do it
• Closed-loop automation includes continuous cycles of awareness, analysis, decision, and execution (AADE)
• Endogenous intelligence embeds AI within network elements, not just in a central cloud
• Single-domain autonomy divides large networks into self-governing units (e.g., RAN or core)
• Cross-domain collaboration institutes horizontal federation across domains
• Multilevel interworking operates consistently across different maturity levels

An open source foundation works with these various characteristics and capabilities, taking business intent and orchestrating AI-driven network optimization at scale within production environments.

The architecture of an autonomous intelligent network

An autonomous intelligent network based on the TM Forum framework relies on a 3-layer architecture supported by a knowledge and intelligence plane, or the artificial intelligence (AI) brain. The business operations layer manages the customer lifecycle and business-level intents, the service operations layer coordinates across domains and manages end-to-end service-level agreement (SLA) assurance, and the resource operations layer includes technology domains including the RAN, core, and edge that run their own closed loops of AADE.

Measurable and real-world impact of autonomous intelligent networks

An autonomous intelligent network can be mapped to the AADE cycle. Consider a 5G core network function experiencing elevated latency:

• Awareness: Monitoring tools detect an anomaly in metrics from the 5G core.
• Analysis: An AI agent queries the knowledge graph for topology context and runs root cause analysis using fine-tuned models.
• Decision: The agent identifies a misconfigured scaling policy and generates a remediation recommendation.
• Execution: Once confidence thresholds are met automation engines execute the remediation playbook—rolling back the policy and verifying that latency returns to SLA compliance.

A similar approach applies to a customer service assistant for network operations. AI leader orchestrator agents can alert subagents to gather customer profiles and configurations. These agents identify outdated firmware or misconfigurations that guide engineers to execute remediation playbooks that fix issues and resolve tickets faster.

Building a strategy for the future

Moving to higher levels of autonomy requires more than just new technology. It requires organizational change. Service providers should start by automating high-volume and low-risk tasks to validate the AADE cycle.

Standardizing on an open source approach across all layers supports interoperability and transparency. It allows service providers to capitalize on community innovation while maintaining control over their technology stack, which is a key component of digital sovereignty.

The Red Hat portfolio provides the operational consistency needed to deploy workloads across on-premise datacenters, distributed edge locations, and hybrid clouds. This foundation prepares service providers for the practical adoption of agentic AI and future technologies like 6G.

Red Hat AI serves as the platform for the knowledge and intelligence plane. It allows service providers to fine-tune open foundation models on proprietary data without compromising data sovereignty. This localized intelligence is critical at the edge for real-time responses to environmental changes.

Red Hat Ansible Automation Platform implements configuration changes with automated playbooks. Red Hat Event-Driven Ansible closes the loop by triggering remediations in minutes, not hours.

Red Hat OpenShift acts as the common cloud infrastructure for autonomous domains (RAN, core, and edge) and supports site intelligence by serving lightweight models at the edge of the network. 

Conclusion

The journey to an autonomous intelligent network is a strategic necessity for the modern service provider. By implementing a pragmatic framework based on open source and AI-driven automation, service providers can reduce operational expenditure while increasing the velocity of new service delivery. Mandating open source interoperability by standardizing open APIs prevents isolated pockets of automation.

Piloting closed-loop use cases that address high-volume and low-risk tasks like automated energy management will prove the value of the automation cycle. This evolution from reactive troubleshooting to proactive intelligent management unlocks new revenue streams and meets growing customer demand.

To find out more, read the following articles and publications: 

• From automation to autonomy overview
• Intelligent framework for autonomous intelligent networks overview
• Maximizing the value of telecommunications automation overview
• Telco autonomous networks choosing the right cloud and framework blog
• AI insights with actionable automation accelerate the journey to autonomous networks blog
• Telco meets AI with autonomous intelligent networks from Red Hat video