For more than a decade, the construction industry had set the target to reduce the utility damage numbers to zero. The sector supported the initiation of the WIBON law (originally called WION) that regulated the utility location and planning information exchange between stakeholders (utility and excavation companies); introduced the utility statutory record exchange system KLIC in the eighties; established the KLO (utility sector board, kabel en leidingoverleg), and took many initiatives to raise risk awareness.
Although it is undeniable that the sector has professionalized and the Netherlands is an international front runner regarding the exchange of utility data, it remains very challenging to identify risks and avoid excavation damages.
The challenge
Over the past five years, the excavation damages numbers have not decreased significantly. Authorities (Agentschap Telecom, and the Ministry of Economic Affairs) have raised their concerns about this since it is likely that the total work volume of excavation work will grow in the upcoming years. It thus becomes even more important to identify and predict excavation damages.
As executive organization of WIBON, and operator of the KLIC-data exchange system, the Kadaster can access excavation process data to analyse, for example, patterns in, and causes of, past excavation incidents. By integrating this information with other geo-data (such as infrastructure maps, maintenance data, information about excavation techniques, historical digging or incident data) it may become possible to better sensitize the industry toward potential threats to excavation damage, or even predict incidents.
The challenge for the sector is to identify what data and which methods can be used to establish a model/platform that could support this analysis and prediction.
This culminates in the following goal of this PDEng, which is to:
Develop a (geo) data-driven model/platform to identify patterns and causes of damage and make predictions about the likelihood that excavation damages occur on construction sites.
Your profile
This project follows a design science approach, in which you are expected to explore the problem context and literature, then iteratively make an innovative design of the data-driven approach, and validate it in a real-life setting. You will primarily focus on integration and analysis of (geo) data, and depending on the needs of the project, may extend this to software prototype development. You will work closely with professionals from Kadaster and spend around 50% of your time at the Kadaster office (Apeldoorn), and 50% at the University of Twente.
Eligible candidates should have a background (completed M.Sc.) in computer science/business and IT or data science (e.g. through possession of certificates, practical experience, or course grades), and/or construction management and engineering/civil engineering and management (digitalization, modelling, data driven approaches). Possession of Dutch language skills is an asset given the work culture at the host company.