Problem Statement

Technical inventory data of traffic infrastructures – e.g. construction plans, maintenance logs or sensor data – are often available in different formats and unclear: as scanned documents or as structured data records in IT systems. Accessing this information requires a lot of manual work and is inefficient. At the moment, specialists have to search for and interpret data from many different sources. There is a lack of an intelligent solution that automatically collates this scattered information and prepares it in a clear and concise manner. The aim is to network different types of inventory data in such a way that they can be used directly for analysis, maintenance and action planning.

Projekt Objective

The HyProTwin project is developing a method to automatically create digital twins of transport infrastructures – using the example of structures on waterways such as locks, weirs and culverts. These digital images of the structures are to intelligently link current and historical data in order to make them more easily accessible and usable. HyProTwin uses a novel combination of two AI methods to achieve this:

• Symbolic AI
  • works with defined rules and logical structures

• Sub-symbolic AI
  • learns patterns from large amounts of data

By combining these approaches, heterogeneous inventory data is to be analysed, linked and made available as a digital twin.

Realisation

The first step is to analyse what information is needed for the maintenance of traffic infrastructures. This analysis forms the basis for the development of an AI-supported search engine that automatically finds data from various sources, structures it and prepares it in a process-orientated manner. With the help of a conceptual information model that describes the meaning and interrelationships of information, a digital twin is created that intelligently links various inventory data and makes it accessible. This enables specialists to find relevant information more quickly and make well-founded decisions. In order to test the practicality of the method, the system is being tested on real waterway structures in collaboration with the Federal Waterways Engineering and Research Institute (BAW). The developed solution will be implemented and validated as a demonstration application.