Another element of network level data collection is GPR. Initially when considering GPR, there is the natural thought that GPR is a "project tool" as opposed to a "network tool", primarily due to the complexities of understanding GPR data. This is further exacerbated by the fact that when you look at the GPR data you are looking back over a long period of pavement history within the deeper layers in the pavement. In contrast the pavement surface history is much younger.
There are simple techniques that can be used over a network. As a first step GPR data can be collected as a background data set. This can be done as GPR collection is quite an affordable data set. Simple automatic measures can then be applied to data such as estimates of the total pavement depth.
As a second stage, GPR layers can be analyzed at a network level. GPR data is intrinsically effective both at small scales (50-100 meters) as well as at large scales (1-2 km). For network level analysis, the data can be analyzed at these larger scales, at an economic rate. At this scale, it is possible to clearly see changes in construction, layering, bottom of asphalt, bottom of base, bottom of sub-base and fill layers. The most useful feature we have found within this domain is to fix the location of any changes in construction between the authorities' pavement management system and the actual change in construction, both with respect to asphalt and base materials.
Currently RPS is also working on methods of subsurface change detection, allowing GPR data to be automatically compared from one year to the next. This allows isolation of areas of sub-surface change. When combined with the surface changed detection methods, the result is a method for detection of sub-surface change with no associated surface change. This drastically reduces the amount of manual analysis required for a road network, reducing it to areas that directly correspond to significant change - outlining areas of moisture ingress, new or changing voids, broken water pipes and seasonal variations.