For most network infrastructure operators, the goal is clear: introduce modern information systems and revise processes to improve efficiency, reduce costs, and improve ROI. Yet, a universal barrier often stands in the way—decades of legacy network documentation. This data, critical to your operations, is often trapped in outdated, inconsistent, and unstructured formats. When data remains locked in AutoCAD files, PDFs, scanned images, and other raster documents, operators lose the opportunity to unlock the true power of modern Network Information Systems (NIS). Automations, advanced search, analytics, network tracing, etc., are simply not possible. 

Although the problem is widespread, it appears in different forms depending on each company’s history, processes, tools, and specific assets. Addressing it requires a structured analytical approach: understanding the nature of the problem, defining the necessary processing steps, selecting and adapting suitable algorithms, identifying typical documentation errors, designing and integrating error-resolution mechanisms, and finally validating the data to ensure high data quality. 

Based on our projects and experiences with major network operators and utilities, we present five distinct, real-world projects that were successfully completed, and show our capabilities to find and deliver tailored solutions for a variety of different situations.  

Case Studies Explained 

1. National Telecom Provider: Legacy unstructured physical network data (paper documents and digital DWG files) converted into a modern NIS establishing a true single system of record. 
2. The Urban Electricity Grid Provider: Existing vector-based electricity grid data cleaned and structured for seamless import into the operator’s NIS. 
3. The Municipal Electricity Utility Provider: The unrefined and inconsistent data that was previously imported from different vector-based databases by the provider was structured, cleaned, and consolidated in the provider’s database, together with the large number of background raster images. 
4. Municipal Gas, Water And Electricity Utility: Unstructured AutoCAD (DWG) vector-based files with embedded scanned raster images converted into structured NIS-ready data. 
5. Energy, Telecommunications And Information Technology Company: Over one million complex and varied network dimensions that remain locked in scanned images were extracted and structured to cancel manual work and prevent immense labor, cost, and error-proneness.

1. National Telecom Provider

Project goal: convert and migrate legacy unstructured physical network data (paper documents and digital DWG files) into a modern NIS to establish a true single system of record. 

Key Project Characteristics 

• Massive data volume: The documentation describing the physical network consisted of paper documents (400,000 images - TIF and JPG files after scanning) and more than 110,000 AutoCAD (DWG) files. A vast number of documents were not georeferenced. 

• Missing and conflicting information: The documentation, including maps, schematic diagrams, splice diagrams, and manhole layouts, contained overlapping and sometimes conflicting representations of the network. These discrepancies had to be carefully identified and resolved. 

• Diverse standards: The documentation lacked geometric and topological consistency. For example, AutoCAD files included: 

• • Lines representing trenches not snapped to points representing manholes, poles, or cabinets. 

• • Duplicate graphical elements layered over each other. 

• • Inconsistent symbols for the same network elements, often without associated attributes. 

• Regional inconsistencies in documentation: The presence of eight regional documentation departments, each applying different standards, further increased the complexity. 

• Existing high-level network inventory in a separate database: Part of the high-level physical network inventory was already maintained in a separate database. Therefore, the converted data had to be aligned and integrated with this existing database. 

Solution Steps: 

• Mass scanning of paper documentation and preparation of workflows for large-scale processing 

• Georeferencing - assigning spatial data to a real-world geographic coordinate system, ensuring that they match actual locations on the Earth.  

• Vectorization of data and creation of digitized links and nodes, along with their properties – newly structured data stored in a local database using our Interactively Assisted Converter (IAC).  

• In a process of vectorization, establishing an in-app ticketing system to address gaps in data and resolve data conflicts collaboratively with the customer in real-time to ensure excellent data quality and consistency.

2. The Urban Electricity Grid Provider

Project goal: Clean and structure existing vector-based electricity grid data for seamless import into the operator’s NIS. 

Key Project Characteristics: 

• Massive data volume : 22,000 vector-based maps in Bentley DGN format. Some of them were outdated. 

• Unstructured graphical data: The trench network was represented with line and point symbols. However, the cable network was partially defined, with graphical elements depicting cable lines and splice closures. These cable lines were not interconnected, resulting in an incomplete graphical representation of the cable network, which had to be reconstructed during the data project.​ 

• Cross-section issues: Cables depicted as dots in cross-sectional diagrams lacked direct relationships to ducts, trenches, and cables shown on the maps.​ 

• The existing high-level network inventory database did not match the vector-based data completely: The cable network data in an existing database was not synchronized with the vector-based DGN maps, and direct automatic correlation was not possible. 

Solution Steps: 

• Efficient file organization: Developed workflows to systematically process 22,000 vector-based maps in DGN format, ensuring streamlined and consistent handling. Each map/DGN file was assigned a status (e.g., not yet processed, in processing, or processed) to track progress. 

• Building geometrically and topologically correct network of trenches and cables. 

• Data stitching: Merged trenches and cables at the borders of each map to create cohesive networks with complete geometrical and topological integrity. 

• Cross-sectional alignment: Digitized cable and duct data from cross-sectional diagrams and linked it to the corresponding trench networks. 

• Data synchronization: Matched and synchronized the physical network data with the existing cable network information database. 

• Interactive validation: Collaboratively resolved ambiguities identified in the data by the customer’s experts using a ticketing system integrated into our Interactively Assisted Converter (IAC) processing application.

3. The Municipal Electricity Provider

Project goal: structure and clean the data imported directly from different vector-based databases for the entire city while simultaneously managing and incorporating continuous, daily updates coming from the field teams without any daily work disruption. Structure a large number of background raster images and import data to the Network Information System (NIS). 

Key Project Characteristics: 

• Massive data volume: 170,000 images and background information to be structured and imported into the GIS-based NIS. 

• Geometric and topological inconsistencies: Following the implementation of a modern NIS, the customer realized that direct import of legacy data resulted in geometric and topological inconsistencies, severely impacting operational efficiency. 

• No interruption of daily operations allowed 

Solution Steps: 

• Structuring of legacy data: Data from various databases and files was converted into a structured format, improving geometry, ensuring complete topological relationships, and properly attributing network elements. 

• Data merging of the newly structured legacy network data that was previously divided into a rectangular grid for the entire city . The outcome was a seamless, structured dataset. 

• Synchronization of physical and pre-existing logical inventories: Using our IAC application, the synchronization between physical and logical network inventories was automated, ensuring accurate matching of elements and the creation of consistent relationships. 

• Quality assurance: All data was validated against a comprehensive set of technical rules defined by the customer. Our IAC application identified and corrected errors, ensuring long-term reliability and trust in the data.

4. Municipal Gas, Water And Electricity Provider

Project Goal: Convert unstructured AutoCAD (DWG) vector-based files that have embedded scanned raster images into structured NIS-ready data. 

Project Characteristics: 

• Unstructured vector-based CAD data: The CAD files had embedded graphics such as lines, points, symbols, and text. While visually appealing, they contained geometric and topological errors. Moreover, the annotations were not linked to the network elements—such as lines representing trenches or cables, and points/symbols representing infrastructure like poles, cabinets, or closures. 

• Raster images – scanned drawings: Thousands of scanned manual network drawings were embedded as raster images in AutoCAD DWG files. These "raster islands" contained data that was incompatible with modern NISs, making the information human-readable only. 

Solution Steps: 

• Raster image vectorization: Converted the embedded, non-vectorized parts of raster images into geometrically and topologically accurate network data. This entirely eliminated raster images, reducing both time and costs associated with maintaining them. 

• Structuring vector-based cad data:  

• • Structured the CAD vector graphics data, ensuring geometric and topological accuracy. 

• • Converted texts into attributes and linked them to the corresponding node or link network elements. 

• • Prepared the data for seamless import into SWA’s modern NIS, effectively eliminating the need for Vector-Based CAD files.

5. Energy, Telecommunications And Information Technology Provider 

Project Goal: A prominent energy company had structured most of its network data, but a critical part remained locked in scanned images: over 1 million dimensions. Manual extraction of these complex and varied dimensions was unfeasible due to the immense labor, cost, and error-proneness. This project was on the critical path to achieve business goals, making the speed and accuracy of the data a crucial requirement. 

Project Characteristics: 

• Manual inefficiency: The process of manually extracting and validating dimensions would be labor-intensive, time-consuming, and error-prone—making it an unfeasible approach. 

• Varied dimension types: The dimensions themselves varied in style, clarity, completeness, and complexity, adding complexity to the extraction process. 

• Tight deadlines: Extracting over one million dimensions within a limited timeframe was crucial, as other network data and system-related projects were scheduled to meet business goals. This project was on the critical path and had to be completed on time. 

Solution Steps:

With our Interactively Assisted Converter (IAC), we implemented an AI-driven, systematic approach: 

• AI-assisted extraction: The AI-powered IAC tool precisely detected and recognized geometric and numerical elements of each dimension. 

• Advanced post-processing: Tailored algorithms were developed to handle diverse dimension types, improving both speed and accuracy in processing the AI-extracted data. 

• Dimension stitching: Ensured completeness by aligning and stitching dimensions at map borders for a seamless dataset. 

• Structuring dimensions: Structured dimensions with properties and ensured they were accurately snapped to both the network infrastructure and surrounding buildings or other man-made structures. 

• Quality assurance: Extracted dimensions were automatically masked within raster images to identify any missing elements for immediate resolution. 

The Winning Solution: Hybrid Approach to Data Conversion And Migration Projects 

A one-size-fits-all approach cannot be used for the projects mentioned above. Such projects require a versatile, powerful processing tool that can adapt to different data types and quality issues, together with process expertise and comprehensive project management. 

Our Interactively Assisted Converter (IAC) platform provides the ideal balance. It functions as an all-in-one data processing solution combining: 

• An AI-driven and rules-based approach for precise data extraction. 

• Automated processing with human-in-the-loop validation to ensure high data accuracy. 

• Advanced processing to georeference data, digitize raster images, stitch maps, and synchronize physical and logical inventories automatically. 

The Results: 

• High-quality structured NIS-ready data  

• Single system of record with automation-ready data  

• Full utilization of NMS accelerates network operations efficiency – faster and higher ROI 

By structuring legacy data, we empower our clients to unlock the full potential of their modern information systems. As one customer noted: 

"With structured data now integrated into our operating support system... we have reduced costs, gained a comprehensive understanding of our network, and feel much more in control." 

Another client highlighted the impact on quality: 

"Our goal was to eliminate the 'garbage in, garbage out' issue, which previously led to costly field surveys and inspections."

You can read more about data migrations and Interactively Assisted Converter here.

Discover how our experts can help you resolve your specific challenges. You can schedule a call with our technical team by filling out the contact form below