By Miha Ušeničnik, Associate Director at DFG CONSULTING
The telecommunications industry is currently undergoing a structural metamorphosis. We are witnessing a transition where telecom automation is shifting from executing predefined, script-based tasks toward the realization of networks that can analyze complex environmental conditions, make real-time decisions, and act with a high degree of autonomy. This evolution—encompassing self-configuration for provisioning, closed-loop optimization, and automated fault recovery—promises to deliver unprecedented levels of operational efficiency and network reliability. Ultimately, this journey aims to enable services to be provisioned, scaled, and adapted with minimal human intervention.
However, as the industry chases the dream of "lights-out" operations, a fundamental hurdle remains: the quality and trustworthiness of the data underpinning these systems. Without a precise, digital "source of truth," the promise of autonomy remains an unreachable horizon.
The State of Play: Mapping the Autonomous Roadmap
The transition toward autonomous networking is already well underway, yet it is currently characterized by a significant "ambition-reality" gap. According to a seminal study by the IBM Institute for Business Value and the TM Forum, titled Navigating autonomous networks, the industry’s strategic intent is clear: 73% of network executives surveyed confirmed that their organizations have developed phased roadmaps to transition toward fully autonomous operations.
Despite this high level of strategic commitment, the ground reality is more modest. Only 6% of Communication Service Providers (CSPs) currently report operating at Level 4, which represents a highly autonomous network state. While the ambition is there, the execution is hampered by legacy complexities. Industry projections suggest that within three years, 22% of CSPs expect to reach that level of maturity. To bridge this 16% gap, the industry must pivot its focus from the software layer alone to the foundational physical data layer.
The Chronology of Discrepancy: Why Networks and Data Diverge
To understand the challenge, one must look at the history of telecom infrastructure. Telecommunications networks have evolved through decades of continuous expansion, technology shifts, mergers, and system migrations. This history has left most operators with a fragmented information architecture.
The Legacy Data Trap
In the early days of network expansion, documentation was often physical or localized. Over time, critical network information has become scattered across a heterogeneous landscape: CAD drawings, Visio files, PDFs, raster images, and disparate spreadsheets, often existing alongside modern central inventory systems.
While these artifacts are invaluable to veteran engineers who know the "hidden" geography of the network, they are essentially "dark data" to modern automation engines. An algorithm cannot parse a legacy PDF to determine the signal path of a fiber optic line.
As-Planned vs. As-Built
A deeper, more systemic issue is the divergence between as-planned and as-built documentation. In the fast-paced environment of field deployments, changes are not always reflected in the system of record. When field teams encounter discrepancies, they often resort to "shadow documentation"—local, unofficial records used to capture the reality of the network. While this allows the network to function in the short term, it creates a persistent, invisible disconnect that prevents automation from ever achieving full reliability.
Supporting Data: The Cost of Inaccuracy
In an automated ecosystem, inaccurate data is not merely a clerical inconvenience; it is a catalyst for operational failure.
Experienced human engineers have a built-in safety mechanism: intuition. They recognize when a record does not align with the physical reality of a site and pause to verify before acting. Automated operations, by contrast, lack this nuanced safeguard. They are literal machines that assume the data fed to them is absolute truth.
The Consequences of "Garbage In"
When an automated system operates on flawed data, the consequences cascade:
- Flawed Path Calculation: If the inventory system suggests a circuit exists where there is a physical break, the automated provisioning system will fail, causing service delays.
- Incorrect Impact Analysis: During an outage, inaccurate connectivity records can lead to the misidentification of affected customers, resulting in misdirected repair crews and unnecessary SLA penalties.
- The "Silent" Asset Problem: Physical assets, unlike software-defined components, cannot report their own state. They cannot self-discover. They require a perfect digital twin. Without an end-to-end, trustworthy representation, the automated "brain" of the network is essentially operating blind.
Intelligent Data Migration: Bridging the Gap
Recognizing the need for trustworthy data is the easy part; the challenge lies in the scale, efficiency, and continuity of the transformation process. Operators are currently sitting on hundreds of thousands of legacy files, while new, often inconsistent documentation arrives daily. Operations cannot be halted for a "data clean-up" project.
The Role of Specialized Transformation
The solution lies in specialized, intelligent migration tools. At DFG CONSULTING, we advocate for an "Interactively Assisted Converter" approach. This methodology moves beyond manual redrawing, which is not only resource-intensive but prone to repeating existing mistakes.
By using AI-driven extraction, operators can take legacy drawings and convert them into machine-readable, structured formats. This process cleanses the data by cross-referencing disparate views—such as spatial maps and splice diagrams—and merging them into a single, cohesive dataset. Where AI encounters ambiguity, human oversight is triggered. This "Human-in-the-Loop" model ensures that edge cases are resolved by experts, while the bulk of the data is processed at a speed and scale that manual efforts simply cannot match.
Implications: The New Role of Visualization
As we move toward higher levels of autonomy, there is a paradoxical need for more—not less—human visibility. As systems make decisions at speeds that defy human comprehension, the risk of "black box" operations increases.
From Documentation to Control
The answer is not to return to manual record-keeping, but to implement dynamic, real-time visualization. Tools such as iNTERACTIVE SCHEMATICS™ allow for the automated generation of network diagrams directly from the same inventory data used by the automation engines.
This creates a vital "Control and Validation" layer. It allows engineers to monitor, verify, and intervene in the decisions made by autonomous systems. Visualization is no longer just for reporting; it has become a necessary interface for human-machine collaboration. It transforms abstract digital data into an intuitive operational view, supporting everything from strategic planning and service provisioning to emergency troubleshooting.
Official Perspective: Trust Before Autonomy
The industry’s push toward full autonomy must be tempered by a commitment to data integrity. Autonomous networks are not merely about the sophistication of the algorithms; they are about the fidelity of the representation of the physical world.
For CSPs, the strategic imperative is clear: you cannot automate what you do not accurately define. The transition to autonomy is a data transformation challenge as much as it is a technology upgrade.
A Call to Assessment
As we look toward the future of connectivity, the defining question for every operator remains: Do we have an effective, trustworthy, and scalable way to convert, validate, reconcile, and visualize the physical network data on which our operations depend?
At DFG CONSULTING, we believe that the first step toward a successful autonomous future is an honest assessment of the present. By analyzing a representative sample of an operator’s data, we can identify quality gaps and transformation opportunities. This diagnostic approach allows operators to build a roadmap that is not just ambitious, but fundamentally viable.
The path to the future is paved with the data of the past. By cleaning, structuring, and visualizing that data today, operators can ensure that their move toward autonomy is built on a foundation of absolute, trusted truth.
About the Author
Miha Ušeničnik has spent more than two decades at the intersection of telecommunications, technology development, and operational transformation. His career includes senior technical and management roles, where he spearheaded broadband network development and contributed to national-level policy, including as the principal author of Slovenia’s Broadband Strategy. At DFG CONSULTING, he leads initiatives to transform complex, legacy network data into structured, automated-ready intelligence.
For further inquiries regarding data assessment or to discuss these strategies in person, please contact [email protected]. We look forward to connecting with industry leaders at Connected Britain 2026, held on 9–10 September.
