Research project

Automated Bridge Defect Recognition

LeanBI offers a platform for automated inspection of bridges based on vision material and on a digital twin. The project will rely on Artificial Intelligence in order to achieve a high damage detection rate, while requiring only a tiny fraction of the volume of annotations used nowadays.

Infrastructure assets, and particularly bridges, require periodic inspection. We aim to reduce the required human factor involvement, mitigate health and safety risks, minimize engineering subjectivity, digitise asset management and enhance sustainable practices in inspection, leading to optimum maintenance strategies.
The solution is a combination of UAV (drone) flights and automated defect detection with AI for inspection and condition assessment of bridges, a globally emerging trend. Inspections and condition assessment will take place on our platform directly on the bridge’s digital twin, drastically increasing efficiency on the inspection (up to 60%) and maintenance (up to 30%) sides and enabling targeted interventions, leading in turn to prolonged life spans (up to 30%) and enhanced sustainability effects. 
We plan to directly utilize raw images taken from the UAVs flights, aiming to automate the defect recognition process without any manual interventions, giving us a USP.

Duration: 01.10.2022 - 30.09.2024

Funding:

Innosuisse

Partner:

OST - Ostschweizer Fachhochschule

HEG-GE - Haute école de gestion

Basler & Hofmann AG