Telecom operators want networks that can configure, optimise, and repair themselves. This shift promises faster service delivery and fewer manual tasks. It also supports more reliable operations as networks grow larger and more complex.
However, autonomous networks depend on one basic requirement: trusted data. If the network record does not match reality, automation can make poor decisions at scale.
A study by the IBM Institute for Business Value and TM Forum shows clear momentum. It found that 73% of surveyed network executives have phased roadmaps for autonomous operations. Yet only 6% of CSPs operate highly autonomous Level 4 network instances today. Within three years, 22% expect to reach that level.
That gap highlights a central challenge for operators. They must improve physical network data before they expand automation. DFG Consulting argues that data quality now sits at the heart of network transformation.
Many telecom networks have evolved over decades. Operators added new routes, upgraded technologies, merged assets, and migrated systems. As a result, key network information often sits across many formats and platforms.
Some data remains locked in CAD drawings, Visio files, PDFs, images, and spreadsheets. Engineers can read these files. Automation systems usually cannot. This creates a major barrier for smarter network operations.
The issue also extends beyond formats. Field teams may change the live network without updating central records. This creates gaps between planned designs and real deployments. Engineers may then create local “shadow documentation” to fill missing information.
For human teams, experience can reduce risk. An engineer may spot a strange record and question it. Automated systems do not always have that instinct. They act on the data they receive.
Physical network assets create another challenge. Cables, ducts, splitters, and passive equipment cannot report their own condition. They cannot confirm where they sit or how they connect. So operators need a reliable digital view of the physical network.
Bad data can cause real operational damage. It may calculate the wrong service path. It may send field teams to the wrong location. During outages, it may identify the wrong customers or services.
Therefore, operators need scalable data conversion and validation. Manual redrawing of legacy files remains common. Yet it takes time and can copy old mistakes into new systems.
Newer tools can extract information from legacy drawings. They can convert it into machine-readable data for inventory systems. AI can help interpret different drawing styles and resolve many quality issues. Human experts can then review unclear or conflicting cases.
Still, humans must remain in control. As automation accelerates decisions, engineers need clear visual tools. Automatically generated schematics can show live network connectivity from one trusted source.
The strategic question is simple: “Do we have an effective, trustworthy and scalable way to convert, validate, reconcile and visualise the physical network data on which network operations depend?”
Autonomous networks will not succeed through intelligence alone. They also need accurate records, clear visibility, and disciplined data governance.
