More precise moisture diagnostics with AI

Client

Institut for Fugtteknik

Industry

Construction

Challenge

Misdiagnosis of moisture damage in buildings

Results

Over 80% diagnostic accuracy

Scope

From research to a scalable diagnostic platform

BCT developed the MoistLab platform together with Institut for Fugtteknik to make expert-based moisture diagnostics more consistent, documentable and scalable.

The project included:

platform and system architecture
data structure and governance
integration of several types of sensor data
integration of AI models and interpretation logic
automatic and standardized reporting
an interactive interface for diagnosis and insight
a foundation for future 3D and AR visualization

The solution

More consistent moisture diagnostics with AI

Moisture damage in buildings has traditionally been assessed based on relatively simple measurements combined with the expert's experience. This can lead to wrong diagnoses and expensive repairs.

MoistLab combines several types of measurements and uses AI to support the assessment, making the diagnosis more precise, consistent and easier to document.

The platform:

combines moisture measurements, thermal images and salt analyses
uses AI to identify the type of moisture and its likely cause
generates standardized diagnostic reports automatically
makes it possible to explore data and results in an interactive interface
is built so 3D and AR visualization can be added later

The goal wasn't to replace the experts, but to give their assessments a stronger and more consistent data foundation.

Process

Expert knowledge turned into a system you can trust

An important part of the work was making expert knowledge understandable, structured and usable in a digital system.

BCT worked closely with Institut for Fugtteknik and their research partners to:

create a robust data structure before AI was added
ensure traceability from the individual measurement to the final diagnosis
build workflows where the results could be understood and verified
present the results so engineers, architects and building owners could all use them

This made it possible to take the solution from a research-based proof of concept to a product that could be used in practice.

Image source: Alexandra Institute

Image source: EHHS

Results

More precise diagnoses and less dependence on subjective assessments

MoistLab has raised both the accuracy and the consistency of moisture diagnostics.

The most important results:

Over 80% diagnostic accuracy. Compared with about 10% before, validated on around 600 inspections

Precise identification of four different types of moisture

Faster and more consistent reporting with less dependence on individual assessments

Better communication of results through interactive visualizations and the option of future 3D and AR display

A scalable foundation for moisture diagnostics that can be used more broadly across the industry

Want to hear more?

MoistLab shows how specialized expert knowledge can be turned into a scalable digital system without losing precision, traceability or trust.

If you also work with complex assessments where expert knowledge, data and consistency are crucial, let's talk.

Talk to us