The pursuit of end to end automation remains a cornerstone of modern telecommunications strategy as operators transition from traditional manual processes to fully autonomous networks. A recent panel by Mobile Europe, moderated by Inderpreet Kaur from Omdia, featuring Afnan Ahmed of Telenor, José Palma of MEO, and Beatriz Ortega of Red Hat explored the current state of this journey and the significant hurdles that remain before the industry can claim true network autonomy.
The discussion cantered on the TM Forum framework for autonomous networks which categorises progress into six levels from zero to five. While many operators are actively automating tasks, the majority currently sit between level one and level two. These early stages involve static or rule based automation where human intervention is required for most decisions. The transition to level three and level four represents a significant leap. At level four, the system manages observability, analysis, and execution with humans only defining the initial intent. This move from deterministic rule based logic to probabilistic reasoning powered by artificial intelligence is where the industry sees both the greatest potential and the most significant cultural resistance.
A recurring theme throughout the session was the challenge of data management. Although telecommunications networks generate vast quantities of information, this data often remains siloed within specific domains like radio access networks or core networks. Creating a unified data mesh or knowledge graph is essential for achieving cross domain automation. The panellists noted that the solution is not simply building a larger data lake. Instead, the focus must be on data correlation. When an issue occurs in one part of the network, the system must understand how that event impacts other domains in real time. Without this level of visibility, end to end automation remains impossible.
Another major obstacle is the sheer complexity of existing operations support systems. Some large operators manage over one thousand different tools, many of which are homegrown or vendor specific. This fragmentation makes it incredibly difficult to implement a cohesive automation strategy. Panellists suggested that operators must radically reimagine their tool suites. The goal should be a vendor agnostic architecture that follows open standards like the Open Digital Architecture from TM Forum. By simplifying the network environment before attempting to automate it, operators can avoid the trap of merely automating existing inefficiencies.
Moving to higher levels of autonomy requires a fundamental shift in how engineers interact with the network. There is a natural fear of losing control when a system begins making its own decisions. To combat this, experts recommend a phased approach where artificial intelligence is first used in an open loop system. In this model, the system provides recommendations that a human operator must validate. Only after the system has proven its reliability over time is the loop closed, allowing the software to execute changes automatically. This process of building trust is vital for ensuring network resilience as systems move toward self healing capabilities.
While operational efficiency and cost reduction are clear drivers, the panel emphasised that autonomy must be viewed as a business transformation rather than just a technical one. The ultimate goal is to enhance customer experience and enable new revenue streams through services like automated network slicing. By achieving level four autonomy, operators can respond to market demands with a speed that manual processes cannot match. This agility is necessary to compete in a digital ecosystem where customers expect near instantaneous service provisioning and seamless performance across diverse network environments.
The shift toward autonomous operations introduces new risks, particularly in cyber security. An automated network could potentially propagate an attack or a misconfiguration much faster than a manual one. There is also the concern of data poisoning, where malicious actors could inject false information to manipulate the decision making process of the network. To mitigate these risks, operators must maintain rigorous data governance and ensure that artificial intelligence decisions remain explainable. Providing a clear audit trail of why a system took a specific action is essential for security and regulatory compliance. Despite these challenges, the consensus remains that the journey toward autonomous networks is an inevitable and necessary evolution for the telecommunications industry.
The video of the discussion as follows:
