The Council for Scientific and Industrial Research (CSIR) has developed a semi-autonomous solution for early pothole detection to potentially replace the often time-consuming expensive manual road inspection.
Showcasing the Visual Surveying Platform (VSP), CSIR senior research technologist Thegaran Naidoo told delegates at the fifth CSIR conference in Pretoria on Friday that the solution was aimed at mitigating the R50-billion-a-year vehicle repair and injury costs resulting from deteriorated roads across South Africa.
Currently piloting within the existing pothole detection workflow of one of the municipalities in Gauteng, the integrated system delivered sensor data collection, visualisation and planning, automatic and manual detection and estimations allowing for a timeous detection and identification of distressed roads and potholes, which could expedite maintenance and repair.
The system comprised a vehicle-mounted mobile sensor equipped with a high-speed camera and a Global Positioning System receiver geotagging and analysing the infrastructure, which was validated by an operator to prevent “false positives” and enhance the accuracy of the information passed on to municipalities.
Existing inspection programmes were dependent on the manual evaluation of road surfaces, which was a time-intensive task.
Automated inspections systems, used as part of a broader inspection programme, could improve road maintenance by providing more in-depth and accurate information on surface properties and road distresses, faster than manual methods.