CONTACT Software is testing how Artificial Intelligence (AI) can significantly accelerate the modernization of German waterways. The solution aims to reduce the effort required to search for information by 60%.
Germany’s waterway infrastructure is in considerable need of modernization, as a look at the age structure reveals: 80% of locks are over 50 years old, with 33% even exceeding 100 years of age (as of 2023). At the same time, there is a lack of experts to carry out the urgently needed measures. One major issue: those responsible spend a lot of time searching for archive information. In the HyProTwin project, CONTACT Software and partners are developing an AI-based solution to relieve their workload and implement projects more efficiently.
Until now, archive data such as construction or maintenance plans for infrastructure buildings had to be painstakingly researched from various sources. The quality of this information varies greatly, from structured data to partially outdated scanned documents. The search accounts for up to 25% of the total project effort. Together with the German Federal Waterways Engineering and Research Institute (BAW) and the Institute for Numerical Methods and Informatics in Civil Engineering (IIB) at the Technical University of Darmstadt, CONTACT aims to reduce this figure to less than 10%.
The partners are developing a hybrid AI approach to interpret archive data and make it searchable. It combines symbolic AI, which works with established rules and logical structures, and subsymbolic AI, which learns patterns from large amounts of data. CONTACT’s CIM Database PLM serves as the project’s data backbone, as it is already successfully used for data and process management by the German Federal Waterways and Shipping Administration (WSV). The AI models are directly integrated into the platform.
HyProTwin automatically consolidates the scattered information into digital twins of the buildings. This enables users to quickly find all relevant documents with a single search query, make informed decisions, and optimize planning – for example, regarding environmental emissions. By rating the search results, they also continuously train the AI models.
Although HyProTwin focuses on waterway structures, it also offers potential for other infrastructure areas. “The approach can also significantly accelerate projects in domains such as road or rail transport,” says Christoph Funk, Senior Consultant at CONTACT Software. The three-year project is funded by the German Federal Ministry of Transport (BMV).
