AI Sensor Technology StoneCrackMon
The Company
Diwo Sculpture Workshop from Paderborn restores monuments and sculptures. An important
part of the work involves the regular inspection of existing objects – because hairline cracks in the stone can endanger stability.
These inspections are necessary but very time-consuming.
Previously, hairline cracks were detected using visual inspections or acoustic tests: experienced sculptors tap and scratch the stone with a hammer and assess its integrity based on the sound. This method is subjective and requires extensive experience – a trained eye and ear. Sculptor Michael Diwo wishes to objectify this process with the help of a system.
The Vision
Making knowledge portable – with a system that makes experience audible: “StoneCrackMon” helps to precisely detect damage in stone – quickly, safely, and independently
of years of experience. It supports not only sculptors but can also be used in monument preservation or construction – wherever accurate crack inspections are critical for safety.
The Solution
StoneCrackMon is an AI-powered sensor system that objectifies and standardizes crack detection. It uses a microphone to analyze the sounds produced by hammer strikes. The evaluation is performed in real-time – the display shows whether a defect is present.
The biggest challenge was translating the expert knowledge of specialists into measurable data. The system had to learn to reliably distinguish between different types of stone, impact intensities, and background noises.