Cracking Statistics

The RPS PaveScout can provide cracking statistics for an entire road network

There are many claims from other crack detection systems, but few provide accurate assessments that represent real cracking results in the field.

At RPS we have conducted comparative assessments, and the results are surprising. The biggest surprise is the quality of the crack maps produced by the PaveScout system.

Very few cracks go unnoticed with the Photometric imaging system - even to the point where we are often detecting the precursors to cracking on a surface.

The result is fixing cracking earlier - with our customers reaping the costs benefits of a longer road life.

As an analysis of our system we compared environmental camera images, detailed non-Photometric surface imaging, and finally Photometric surface imaging. The ground truth of our analysis was conducted by using manual crack mapping (walking the road in the field).

The right section of the image shows the original surface condition. By brushing the surface the cracking can be revealed. Much of this cracking was missed in the manual mark-up of the road surface. It was only after the detection of the cracking with the photometric image analysis did we later go back and detect this cracking in the field.

Crack detection

Spray sealed road with debris in the cracks

For Asphalted Surfaces:

For asphalted surface the crack statics is as follows:

Crack Width (mm) Asset Camera W/O Photometric With Photometric
0-1 0% 29% 50%
1-2 12% 67% 81%
2-3 13% 100% 100%
3-4 26% 96% 100%
4-5 17% 100% 100%
>5 0% 100% 100%
False + 4% 30% 3%
Missed in manual survey 0% 3% 3%

As can be seen, high resolution surface images of the pavement with artificial lighting is not the only factor that is required to detect cracks in a surface.  This is due to a large amount of crack-like features that naturally appear in an asphalt surface from vehicle use.  The result was a high false positive rate.  In contrast the Photometric surface imaging system not only increases the amount of detected cracks, but also reduces the amount of false positives.

For Spray Seal Surface:

For spray seal or chip seal surfaces, the statistics are as follows.  The roads used in this study were not freshly cleared chip seal surfaces; instead they were surfaces with significant loose rock debris.

Crack Width (mm) Asset Camera W/O Photometric With Photometric
0-1 0% 25% 38%
1-2 0% 13% 50%
2-3 0% 13% 63%
3-4 0% 43% 86%
4-5 8% 31% 69%
>5 14% 10% 86%
False + 4% 40% 11%
Missed in manual survey 0% 3% 142%

The crack detection rate was worse for spray seal surface in contrast to asphalt surface, highlighting the difficulty of detecting cracks in this type of surface.  As with asphalt surface, the false positive rate went down with the photometric imaging system, however the biggest surprise was the amount of cracks detected by the system, that were not detected in the field.  As in the Photometric imaging system was able distinctly see the fine debris in the cracks and distinguish this from the coursed debris of the road.  After the analysis we went back into the field, and swept the road.  To our surprise the cracks detected by the Photometric system appeared.  The result was that the Photometric system was better at detecting cracks in the field when compared to manual assessment.

Return to top