Extracting meta data from construction plans

City of Venlo


City of Venlo



The city of Venlo in the Netherlands digitized over 200,000 construction plans of all kinds of buildings and other structures in their region. These plans date back decades. Some of them even date back to the early years of the 20th century, so a considerable part of the plans are still drawn and annotated by hand. In the context of the Environment Act, the city of Venlo wants to be able to retrieve construction plans based on location or other data. This means that the building plans must be analyzed and the required metadata made explicit.

If the execution of a task requires human intelligence, but the handling of the task is inhumanly large, then the deployment of artificial intelligence is valuable. Analyzing construction plans is a task that requires human intelligence. The number of over 200,000 plans, however, makes it undesirable for a person to perform that task. As an example – with an average processing time of 10 minuten per plan – it would take a group of 10 people well over a year to process all the construction plans.

With our Geonizer™ platform we are able to process the construction plans and solve this problem for Venlo.  The intelligent agents in Geonizer convert images containing text into processable text. After that, the agents analyze the textual information looking references to locations and for the required metadata. The extracted data is validated and presented back to our client. Problem solved.