In an era where telecom networks are becoming increasingly complex, one company has set a remarkable benchmark by achieving a level of autonomy that seemed unattainable just a few years ago, marking a significant milestone in the industry. China Mobile, recognized as the world’s largest 5G network operator by subscriber count, has successfully attained Level 4 Autonomous Networks (AN) in its Network Operation Centers (NOCs). With a staggering base of over 1 billion mobile subscribers, including 578 million on 5G, the scale of this accomplishment cannot be overstated. The transition to highly autonomous systems, often referred to as “dark NOCs,” minimizes human intervention and redefines operational efficiency in an industry burdened by manual processes. This milestone not only highlights the potential of cutting-edge technology but also positions China Mobile as a pioneer in telecom innovation, addressing the pressing need for scalable and reliable network management.
The journey to this point involved overcoming significant hurdles inherent in managing such a vast infrastructure. By integrating advanced artificial intelligence (AI), including telecom-specific Generative AI (GenAI), China Mobile has automated critical functions like fault management and customer complaint resolution. The deployment of intelligent agents and co-pilots, supported by robust AI models, has led to faster response times and fewer errors, directly benefiting both operational workflows and customer satisfaction. This achievement, measured against the TM Forum’s AN Levels assessment, showcases a strategic blend of technology and vision, setting a new standard for what telecom operators can achieve with automation.
The Path to Autonomy
Defining Level 4 and Overcoming Challenges
The shift to Level 4 Autonomous Networks represents a profound transformation for China Mobile, moving from labor-intensive, error-prone manual processes to near-complete autonomous decision-making. In earlier systems, tasks such as IP backhaul activation required a cumbersome 54 steps, with 41 of those needing human input. This often resulted in prolonged delays and frequent mistakes, undermining efficiency across the network. Achieving Level 4, as defined by industry standards, meant automating these intricate processes to eliminate human oversight in routine operations. The focus was on creating a seamless system where decisions are made by intelligent tools, tackling inefficiencies head-on. This transition addressed not only operational bottlenecks but also the high costs associated with manual labor, paving the way for a more agile and responsive network infrastructure.
Beyond the technical leap, the challenge lay in redefining workflows that had long relied on human intervention. Fault resolution, for instance, previously demanded an average of 45 minutes per ticket in backend operations, with total resolution times stretching up to 100 minutes due to on-site coordination. By prioritizing high-value scenarios across domains like radio access networks (RAN) and core networks, China Mobile systematically reduced these delays. The result was a fundamental overhaul of how network issues are identified and resolved, ensuring that automation could handle complex tasks with precision. This shift not only improved response times but also set a precedent for how large-scale telecom operators can rethink traditional operational models to meet modern demands.
Pilot Success Stories
Initial steps toward this ambitious goal were tested through pilot projects in Guangdong and Zhejiang provinces, regions with distinct operational challenges and high subscriber density. In Guangdong, the focus centered on IP backhaul faults and home broadband complaints, yielding impressive early results. Intelligent agents took on 30% of backend roles for IP faults, while mean time to repair (MTTR) for broadband issues was slashed by 50%. These outcomes demonstrated the potential of AI-driven automation to transform critical pain points into streamlined processes. The success in this province, China’s most populated, provided a strong foundation for broader application, proving that localized solutions could address specific regional needs effectively.
Meanwhile, Zhejiang province showcased equally compelling results, particularly in RAN and core network fault management. Digital employees, powered by AI, handled up to 40% of backend operations, achieving an 87% reduction in core network fault MTTR. Additionally, a 75% complaint pre-emption rate was recorded, meaning issues were often resolved before customers even reported them. These pilot successes validated the approach, offering concrete evidence that automation could deliver measurable improvements in both efficiency and service quality. The lessons learned from these trials were instrumental in refining strategies for a national rollout, ensuring that the solutions were robust enough to handle diverse operational environments across the country.
Technology Behind the Transformation
AI Innovation and GenAI Integration
Central to China Mobile’s groundbreaking achievement is the deployment of sophisticated AI technologies, particularly telecom-specific Generative AI (GenAI), which powers intelligent agents and co-pilots. These tools, built on models boasting over 10 billion parameters, have achieved an impressive accuracy rate of up to 97% for operational queries. Techniques like Retrieval-Augmented Generation (RAG) played a crucial role in minimizing inaccuracies, often termed “hallucinations,” reducing them to a mere 3%. This high level of precision ensures that automated systems can handle intricate tasks such as fault diagnosis and customer issue resolution without the need for constant human oversight. The integration of such advanced AI marks a significant departure from traditional methods, enabling a level of autonomy that redefines network management.
Further enhancing this technological leap is the focus on natural language processing and intelligent Q&A capabilities embedded within these AI agents. Role-based co-pilots and scenario-based agents are designed to address specific operational needs, from troubleshooting network faults to responding to customer inquiries. This tailored approach allows for rapid decision-making, cutting down response times that once hindered service delivery. By leveraging GenAI, China Mobile has created a system where automation not only replicates human tasks but often surpasses them in speed and accuracy. The impact of this innovation extends beyond internal operations, directly enhancing the reliability and responsiveness that subscribers experience daily.
Model Synergy
A key element of China Mobile’s strategy lies in the strategic combination of large and small AI models to achieve a balance between generalization and specialization. Large models, with their extensive parameters, provide a broad understanding of network dynamics, enabling them to address overarching trends and complex scenarios. Smaller, more specialized models focus on niche tasks, ensuring precision in targeted areas like specific fault types or regional data patterns. This synergy creates a closed-loop automation system where processes are not only initiated but also completed autonomously, minimizing the need for manual intervention at any stage. The result is a highly efficient operation that can adapt to a wide range of challenges without sacrificing accuracy.
Another innovative aspect of this approach is the adoption of no-code development platforms, which allow non-technical staff to contribute expertise through natural language inputs. This democratization of input ensures that domain knowledge from various teams is integrated into the automation process, enhancing the system’s effectiveness. By bridging the gap between technical and operational teams, China Mobile has created a collaborative environment where AI tools are continuously refined based on real-world feedback. This model synergy not only supports current operational needs but also builds a flexible framework that can evolve with future technological advancements, ensuring long-term sustainability in an ever-changing telecom landscape.
Strategic Implementation and Industry Standards
Scaling Across Regions
Implementing Level 4 autonomy across a network serving over 1 billion users required a meticulous, bottom-up strategy that prioritized localized adaptation. China Mobile recognized early on that a one-size-fits-all approach would not suffice given the diverse business models and data patterns across its provincial subsidiaries. Starting with pilot regions like Guangdong and Zhejiang, AI models were trained on local data to ensure relevance and accuracy in addressing specific operational challenges. This tailored training allowed the automation systems to handle regional nuances effectively, from varying subscriber behaviors to unique network configurations. The success in these initial areas provided critical insights that shaped a scalable model for nationwide deployment.
The process of scaling, however, was not without its complexities. National rollout demanded rigorous testing to ensure that solutions remained effective across different environments, where data patterns could differ significantly. By focusing on a gradual expansion, China Mobile mitigated risks associated with rapid implementation, ensuring that each region benefited from refined AI tools adapted to its needs. This methodical approach also allowed for continuous feedback loops, where performance metrics from one area informed adjustments in others. The result was a cohesive national system that maintained high standards of autonomy and efficiency, regardless of geographic or operational variations, solidifying the company’s ability to manage its vast infrastructure with precision.
Leveraging TM Forum Frameworks
A cornerstone of China Mobile’s success in achieving Level 4 autonomy was the strategic use of industry-standard frameworks provided by the TM Forum. Assets such as the Business Process Framework (eTOM), ODA Functional Framework, and AN Journey methodology offered structured guidance throughout the automation journey. These tools helped define clear business requirements, assess existing gaps in operations, and design targeted solutions to bridge those gaps. By adhering to such standardized approaches, the company ensured consistency in implementation, while also facilitating seamless collaboration with technology partners. This alignment with proven frameworks was instrumental in transforming complex operational goals into actionable, measurable outcomes.
Equally important was the role these frameworks played in fostering a unified understanding across teams and stakeholders. The TM Forum’s methodologies provided a common language that streamlined communication, ensuring that every phase—from planning to execution—was aligned with industry best practices. Beyond internal benefits, China Mobile’s engagement with these standards extended to contributing back to the TM Forum, particularly in areas like GenAI agent designs for telecom applications. This commitment to shared progress reflects a broader vision of advancing industry practices, ensuring that the lessons learned from this milestone benefit other operators striving for similar levels of automation and efficiency.
Measurable Impacts and Future Implications
Operational and Financial Gains
The tangible outcomes of China Mobile’s shift to Level 4 Autonomous Networks are evident in the dramatic improvements in operational efficiency and financial savings. By the end of last year, automation had effectively replaced the equivalent of 5,500 full-time roles, reducing backend manpower needs by 30%. Additionally, frontline installation and maintenance staff saw over 5% in labor savings, reflecting a significant cut in operational costs. Mean time to repair (MTTR) for faults and complaints dropped by an average of 30%, with some regions achieving reductions as high as 87%. These metrics highlight how automation has streamlined critical functions, allowing for faster, more reliable network management while slashing expenses that once burdened the company’s bottom line.
Financially, the implications extend beyond mere cost reductions to a broader strategic advantage in a competitive industry. The savings achieved through manpower reductions and process efficiencies have freed up resources that can be reinvested into innovation and infrastructure upgrades. This positions China Mobile to stay ahead in the rapidly evolving telecom sector, where staying competitive often hinges on balancing cost control with service quality. Operationally, the near-elimination of human intervention in routine tasks has minimized error rates, ensuring that network uptime and reliability are maintained at unprecedented levels. These gains collectively underscore the economic and practical benefits of embracing advanced automation at scale.
Customer-Centric Outcomes
From a customer perspective, the impact of China Mobile’s automation efforts has been transformative, particularly in enhancing service reliability and responsiveness. Faster fault resolutions, driven by a 30% average reduction in MTTR, mean subscribers experience less downtime and quicker fixes to issues that once lingered. In certain areas, proactive fault identification has reached a 75% complaint pre-emption rate, resolving potential problems before they even reach the customer. This forward-thinking approach not only reduces frustration but also builds trust in the network’s dependability, setting a high benchmark for what telecom service should be in terms of speed and consistency.
Moreover, the focus on customer complaint handling through AI-driven agents has streamlined interactions, ensuring that issues are addressed with precision and efficiency. Subscribers benefit from a network that anticipates and mitigates disruptions, often without any need for manual reporting. This elevation in customer experience reflects a deeper shift in how telecom operators can leverage technology to prioritize end-user satisfaction. As other industry players take note, the standard set by these outcomes suggests a future where customer-centric automation becomes a core competitive differentiator. Looking ahead, the challenge will be to sustain and expand these improvements, ensuring that every subscriber, regardless of region, experiences the full benefits of an autonomous network.
