EngD position: Data-Driven Reliability Analysis of Underground Infrastructure Maps (ZoARG EngD)

Background

The Netherlands operates a digital system for exchanging location data of infrastructure networks, such as underground cables and pipelines. This system (known as KLIC) is managed by Kadaster, while network owners are responsible for providing accurate data about the location of their buried assets.

Many utility maps used in construction projects combine outdated and recently surveyed network drawings. As a result, it is often unclear whether a map is up-to-date, complete, or accurate. To verify this, project teams dig trial trenches during construction and maintenance works to compare the actual underground situation with the map.

The Challenge
Contractors and municipalities hold hundreds of reports containing sketches and analyses of trial trenches. These often include comparisons between the actual cable positions and the official network maps (KLIC deliveries). A analysis of these datasets yield valuable insights into the conditions under which maps are more or less reliable. However, such an analysis has not yet been conducted. This leaves stakeholders unable to statistically and systematically assess the quality of official network maps.

Could this be done differently? Is it possible to systematically analyse all trench documentation using data-driven methods to better interpret the reliability of underground maps?

Your Role
As an EngD candidate, you will work under supervision to develop a data-driven model for assessing the reliability of official network maps (KLIC deliveries). You will build a prototype algorithm that can be used by infrastructure and civil engineering professionals to better evaluate cable and pipeline location data.

Your tasks will include:

1. Analysing existing trench datasets (content, format, resolution, completeness)
2. Defining the concept of ‘reliability’
3. Exploring suitable data mining and analysis methods
4. Collecting case data
5. Selecting predictive parameters
6. Training algorithms (e.g. decision trees, gradient boosting, neural networks)
7. Validating the prototype with end users

Your Team: ZoARG
This project is part of the ZoARG2.0 programme, in which seven major excavation chain partners (Alliander, Enexis, Gasunie, Heijmans, Kadaster, KPN, Siers, and Vitens) collaborate on research, development, and education in excavation damage prevention and responsible digging. Three of these organisations form the project steering committee.

Between 2025 and 2028, five EngD projects will be launched within ZoARG2.0. Candidates will work with UT researchers to advance scientific knowledge (digital technologies, process models, and training) aimed at improving safety in the excavation sector. ZoARG projects are carried out in close collaboration with industry partners, allowing candidates to develop both academically and professionally.

Start date: March 2026
Duration: 2 years