Case Study: BAM Infra Nederland achieves faster, more accurate road damage detection with Microsoft Azure Machine Learning

Construction company uses machine learning solution to rapidly identify, plan, and repair damaged roads

BAM Infra Nederland, a Dutch infrastructure company that designs, builds, finances, and maintains asphalt roads, needed a faster, less labor-intensive way to inspect road damage. Working with ICT Group, it turned to Microsoft Corporation and Azure Machine Learning to automate the analysis of photos and videos of asphalt, replacing a slow manual inspection process.

Microsoft Corporation helped BAM Infra Nederland implement an Azure-based machine learning solution that identifies damaged roads and classifies the type of damage. The system achieved nearly 100% accuracy for detecting road damage and up to 75% accuracy for identifying damage type, enabling inspectors to review only selected images and reducing work from days or weeks to hours.

BAM Infra Nederland

Kitting Lee

Business and Commercial Manager