Optimize inspections of road assets with AI

HAL24K’s Digital Inspection of Roads solution augments and digitalizes your daily road inspections using video footage from consumer-grade cameras. This is possible through the powerful combination of machine learning models to automatically identify assets and detect defects, and workflow software to easily view, correct, complement and share the results.

This helps your organization gain continuous insight in the state of road infrastructure, optimize reporting and maintenance planning.

How does DIGITAL INSPECTION work?

The process

Digital Inspection of Roads serves as the ultimate co-pilot for inspectors as they determine the state of their road infrastructure. They can safely stay in their car while their camera records the road and its assets, and can be confident that the solution will cover detections such as road signs, marker posts, asphalt damage and fading markings. Videos can also be recorded by colleagues and commuters, with inspectors receiving the feeds in one central dashboard.

As videos are ingested, inspectors select the relevant machine learning models to detect assets and anomalies. All detections are plotted on a map and viewed as downloadable lists. With a single mouse click, individual detections can then instantly be visualized by the specific video frame and downloaded.

The results can easily be improved as inspectors update or add to each inspection round by manually labeling items of interest in the solution’s annotation tool. This completes the output of each inspection, and also feeds new data back to the operational machine learning models for continued learning to improve their accuracy and efficiency.

As the video archive builds up, so does your understanding of your road infrastructure, the quality of your inspections, and the ability to demonstrate compliance and improve stakeholder relationships.

The platform

Based on HAL24K’s Computer Vision Workbench – a fully integrated automated workflow module within Dimension – the Digital Inspection of Roads solution offers video preprocessing and pipeline management, annotation tools, auto-annotation functionality for one or multiple machine learning models, and comprehensive reporting functionality.

Digital Inspection of Roads is robust, secure and open by design, enabling integration with third-party systems such as asset management, GIS, or CAD systems.

Teams can now efficiently track and report on road conditions, plan maintenance and optimize workflow.