Guidance

Section 2: measuring surface condition using automated visual methods

Published 17 December 2024

Applies to England

TRACS

TRACS (TRAffic-speed Condition Surveys) were developed to provide a consistent method of measuring the surface condition of trunk roads using an automated survey machine that would not be disruptive to traffic. TRACS surveys have been carried out for National Highways (NH) and its predecessors on the trunk road network since 2000. The surveys are commissioned centrally by NH (formerly Highways England) and carried out by a single independent contractor using a survey vehicle specifically constructed for the purpose of undertaking the surveys, shown in Figure 1.

The TRACS survey employs laser-based methods to measure the shape and texture of the road surface at traffic-speed. Downward facing digital video equipment is used to collect pavement images which are used to measure cracking. Forward facing video is also collected. The survey employs inertially assisted differential GPS combined with distance measurement equipment, to provide accurate location referencing.

Figure 1: The TRACS survey vehicle (RST27, used since 2006)

Since 2017, TRACS surveys of the trunk road network are undertaken:

• annually on all lanes of mainline carriageways  

• annually on lane 1 on slip roads

TRACS surveys are subject to a detailed quality assurance regime, which is supervised by an independent auditor. The TRACS survey vehicle is subject to accreditation testing before being given a certificate to commence surveys. The system is then re-accredited every 3 months. Regular contractor’s repeat surveys are carried out, where repeat surveys are undertaken by the contractor and checked for consistency. Contractors also provide data from randomly selected sites for comparison with the results of surveys undertaken by an independent survey vehicle.

Following processing of the raw data, TRACS data is delivered to National Highways in the form of Base Condition Data, a bespoke format that has been developed to enable the data to be loaded into the National Highways’ Pavement Management System, (HAPMS). The TRACS data is used in a range of applications, including:

• identification of lengths in need of further investigation

• support for SCRIM site investigations

• value management (scheme prioritisation)

• scheme design

• calculation of National Highways’ Key Performance Measure (KPM)

Guidance on the use of TRACS data is provided in the England National Application Annex to volume CS 230 of the Design Manual for Roads and Bridges (see below).      

SCANNER

The SCANNER (Surface Condition Assessment for the National Network of Roads) survey was initially introduced as a traffic-speed survey of the principal road network under the name TRACS-Type Survey (TTS). As the name suggests, TTS was based on the Trunk Road TRACS survey, to provide a consistent method of measuring the surface condition of road carriageways, using automated road condition survey machines, throughout the United Kingdom.

TTS became the recommended survey method for local authorities to use for the calculation of the BV96 indicator on ‘A’ class roads in England in 2003 to 2004. The TTS survey later became known as SCANNER and its application expanded from ‘A’ roads to all classified roads. SCANNER surveys are now mandatory for the assessment of the classified road network.

In England, the SCANNER Road Condition Indicator (RCI) is the only method used for the calculation of performance indicators for the reporting of condition of the classified road network, initially as BV223 and BV224a, and (from 2008 to 2009) as National Indicators NI168 (‘A’ roads) and NI169 (other classified roads). As of 2011, the National Indicator set was abolished and replaced by the Single Data List where the former National Indicators 168 and 169 are listed as 130-01 and 130-02 respectively.

SCANNER surveys in England are commissioned by local authorities either individually or through consortia. There are currently two survey contractors operating SCANNER accredited survey vehicles, of which typical examples are shown in Figure 2. The measurement equipment installed on SCANNER survey vehicles is fundamentally similar to TRACS (described above), but there are technical differences, as a result of the requirement for different reported parameters, described later in this document.

Figure 2: Typical SCANNER survey vehicles

The Department for Transport (DfT) specifies that SCANNER surveys of the local road network should be undertaken:

• annually, covering on average at least 45% of the ‘A’ road network in both directions

• annually, covering on average at least 42.5% of the ‘B’ road network in both directions

• annually, covering on average at least 40% of the ‘C’ road network in one direction

Whilst roundabouts and slip roads may be covered by the survey vehicle during the process of driving over the network, the data from these lengths are not required as part of the survey and are not included in calculations of RCI or coverage.

All these figures are minimum averages for the most recent two years of surveying. For example, an ‘A’ road figure reported for 2014 will use data gathered between April 2012 and March 2014 covering 90% of the ‘A’ road network in both directions.

However, some local authorities undertake surveys to cover a greater proportion of the network each year, either covering more than the minimum length on classified roads (in effect increasing the frequency of survey) or including some unclassified roads (particularly those carrying more traffic, where the carriageway may be in a higher category in the maintenance hierarchy).

SCANNER surveys are subject to a detailed quality assurance regime, which is supervised by an independent auditor. Each SCANNER survey vehicle is subject to accreditation testing before being given a certificate to commence surveys. The system is then re-accredited annually. Regular contractor’s repeat surveys are carried out, where repeat surveys are undertaken by the contractor and checked for consistency. Contractors also provide data from randomly selected sites for comparison with the results of surveys undertaken by an independent survey vehicle.

Following processing of the raw data, SCANNER survey data are delivered as an HMDIF file, the defined standard format for loading survey data into a UKPMS compliant pavement management system. A key use of UKPMS with SCANNER data is in the processing to produce the SCANNER Road Condition Indicator (RCI). The SCANNER data and the RCI are used in a range of applications, including:

• identification of schemes

• prioritisation of maintenance

• support for SCRIM site investigations

• calculation of Single Data List items 130-01 and 130-02

The SCANNER specification defines the technical requirements for SCANNER survey parameters, accreditation testing, and quality assurance.

In recent years, some Local Authorities have moved to alternative surveying methods and technologies for the purposes of 130-01 and 130-02 calculation. The comparability of such data is currently untested; entries that have been calculated using alternative means are highlighted in tables where appropriate.

The DfT is currently engaged in implementing the new BSI standard PAS 2161 published in September 2024. DfT produced the new standard alongside BSI, TRL and the wider industry, they are now working towards ensuring quality and comparable of data from a range of survey technologies, and compliance with the new standard.

Consistency of SCANNER data

SCANNER surveys are automated surveys carried out using accredited SCANNER survey vehicles. Local authorities commission them to assess the surface condition of their classified ‘A’, ‘B’ and ‘C’ road network - further detail can found later in this document.

Fleet consistency (cracking) – Road Conditions in England 2015

Following the validation process carried out by the department for the Road Conditions in England publication 2015, further investigations were undertaken on the SCANNER survey data. This was completed with the assistance of TRL. TRL are quality auditors for automated condition measurement under the UK SCANNER survey, for the whole of the UK classified road network.

It was found that a particular survey vehicle (ARAN26, which ceased undertaking SCANNER surveys at the end of 2014 to 2015 survey year) that was used by certain authorities in different years, reported relatively higher levels of cracking (defined later in this document) when compared against the fleet of accredited SCANNER vehicles, leading to a higher RCI score.

This meant that, although the vehicle satisfied all the relevant acceptance process tests to meet the specification required, its tendency to detect a higher proportion of cracks present on the road surface in comparison with other vehicles in the fleet may have led to a degree of bias being introduced into the assessment of cracking. This in turn could affect the reporting for the authorities where this vehicle was used. All the figures that could have been affected have been marked in the table accordingly.

Fleet consistency (cracking) – Road Conditions in England 2016

Following the validation process carried out by the department for the Road Conditions in England to March 2016 publication, further investigations were undertaken on the SCANNER survey data. This again was completed with the assistance of TRL as the quality auditors for automated condition measurement under the UK SCANNER survey.

TRL confirmed that, for the 2014 to 2015 survey, the measurement of cracking reported by the ARAN1 and ARAN2 survey vehicles lay slightly outside the other vehicles in the SCANNER survey fleet. This meant that, although these vehicles satisfied all the relevant acceptance tests to meet the specification requirements, they had a tendency to report moderately higher levels of cracking in comparison to the levels reported in 2013 to 2014 and in 2015 to 2016, and hence moderately higher levels than other SCANNER vehicles.

This change was small, and not as significant as seen with ARAN26. However, Authorities for whom surveys were undertaken in 2014 to 2015 using ARAN1 or ARAN2 could see a small drop in the RCI where a different survey vehicle was used in the following year (2015 to 2016).

Fleet consistency (texture) – Road Conditions in England 2016

Until 2016, TRL required that a calibration correction be applied to the texture (SMTD) data provided by the Tempest vehicles, to fully meet the specification requirements and optimise the consistency of the SCANNER survey fleet for the measurement of Texture. As a result of ongoing improvements to Tempest, the need for this correction was removed from the start of the 2016 to 2017 survey.

As part of the ongoing QA checks that are applied throughout the survey, TRL identified some datasets where the calibration had not been applied in the 2016 survey. This was tracked to a processing issue, and the data was re-processed (by the SCANNER contractor) and corrected data delivered to all the authorities identified.

However, there is a possibility that this problem may have affected a small number of authorities in the previous year’s (2014 to 2015) survey. As a result, the SCANNER data from these authorities could exhibit a small increase in the RCI. It has not been possible to identify, in the table, the individual local authorities that were affected.

Fleet consistency (cracking) – Road Conditions in England to March 2018

As a result of ongoing improvements, a new crack detection system was installed on the Tempest 3 device during 2017. The device was tested against the requirements for SCANNER surveys. Although the device satisfied the relevant acceptance tests, it was found that the device tended to report higher levels of cracking in comparison to other SCANNER vehicles. 

With the objective of optimising the consistency of the SCANNER fleet, an attempt was made to correct for this. This delivered improved consistency, whilst still satisfying the relevant acceptance tests. However, because of this work, the levels of cracking associated with this vehicle are now moderately lower in comparison to other SCANNER vehicles. Therefore, the SCANNER data from these authorities could exhibit a small decrease in the RCI. It has not been possible to identify all of the individual local authorities that are affected.

Research was commissioned in 2016 to investigate fleet consistency problems with the measurement of cracking. The research proposed methods for testing consistency during the accreditation programme, including better identification of outliers and tightening of the range of measurements reported by the SCANNER fleet. These assisted in managing, and minimising, the consistency issues identified with one vehicle in 2017, and work is ongoing to determine how further consistency improvements could be achieved.

Fleet consistency (cracking and rutting) – Road Conditions in England to March 2019

Building on the recommendations of the research commissioned in 2016 to investigate questions over fleet consistency in the measurement of cracking, trials were carried out in the summer of 2018 to compare the SCANNER crack measurement systems on common sites. Based on the outcomes of the trials, work was undertaken on the Tempest 1 crack detection system to better align its crack detection with the fleet. The improvements to Tempest 1 slightly reduced the levels of cracking reported by the device. However, the QA process has not identified any significant effect on the RCI for authorities surveyed using this device in 2018 to 2019.

No changes have been made to Tempest 3’s cracking system in 2018 to 2019. Therefore, Tempest 3 continues to report slightly lower levels of cracking in comparison to the fleet. This may continue to have a small effect on the RCI for authorities surveyed using this device. However, it has not been possible to link this to changes to the RCI in specific authorities.

A new survey vehicle has been accredited in 2018 to 2019 – RAV16, which employs the latest technology to deliver a wider measurement width, with improved removal of features that may otherwise adversely affect the measurement of rutting. The new system has been found to report slightly lower levels of rutting, mainly on principal roads. The 2018 to 2019 SCANNER QA audit was not able to clearly link changes in rut measurement on changes to the RCI in specific authorities.

Fleet consistency – Road Conditions in England April 2019 to March 2021

This period saw the introduction of 2 new survey vehicles, Tempest 4 by Ginger Lehmann (now part of XAIS-PTS) in 2019 to 2020 and MFV1 by PTS in 2020 to 2021.

It was found that the new PTS MFV1 device reported slightly lower levels of cracking than the fleet on some surface types. As the lower level of cracking was linked to only certain surfaces, the 2020 to 2021 SCANNER QA audit was not able to clearly link this behaviour to changes to the RCI in specific authorities. Furthermore, the device surveyed only a low number of authorities in 2020 to 2021.

As part of continuous improvements to SCANNER fleet consistency, the auditor worked with Ginger Lehmann in 2020 to 2021 to improve the consistency (against the rest of the fleet) of the Tempest 1 and Tempest 4 rut measurements. Work was carried out to reduce the tendency of these devices to report proportionately lower levels of rutting on some road shapes. This resulted in a slight increase in the depths of rutting reported in some surveys, but also resulted in improved fleet consistency.

As the focus was on measurements carried out on surfaces with relatively low depths of rutting, it is expected that the changes would mainly affect the values reported on lengths with depths of rutting falling below the lower RCI threshold, and therefore they should not have a significant effect on the RCIs reported in, and beyond, 2020 to 2021.

In 2020 to 2021, WDM implemented changes to calculation of SMTD texture measurements for their RAV devices. These changes aimed to reduce the adverse effects of “spikes” present in the raw texture laser measurements. The changes resulted in a small increase in the level of SMTD reported by these vehicles. However, this was also associated with an improvement in the consistency of the measurement of SMTD across the fleet (for example, between survey providers). It was found that these changes resulted in a small decrease in the RCI. This is likely to have had a small effect on the RCI for several authorities, as a result of the size of the RAV fleet. However, the 2020 to 2021 SCANNER QA audit was not able to firmly quantify the significance of this for each authority. It was considered that the consistency benefits offered by the processing changes outweighed the small changes in the RCI. Therefore, the processing changes were retained for future surveys.

Further information on this issue can be provided by the SCANNER Auditor (TRL), [email protected].

Defects measured by both TRACS and SCANNER, and used within condition indices and indicators – the core parameters

The TRACS and SCANNER surveys provide several parameters describing the condition of the pavement surface. Currently, some of these could be considered the “core” parameters that are used in condition indices such as the SCANNER RCI. There are further measures, some of which have been developed more recently, that can be used by highway engineers to assist in targeting lengths for further investigation. As more experience is gained with these measures, they may be included within condition indices in the future.

Transverse profile – Rut Depth

TRACS rut depths are determined from Transverse Profile data, recorded at 100mm intervals over a width of 3.2m. The general principle of the rut depth calculation is to replicate the use of a 2m straight edge with one end positioned close to the left edge of the transverse profile for the nearside rut, and close to the right edge of the transverse profile for the offside rut, as shown in the diagram below. Rut depths are measured perpendicular to the straight edge (Figure 3).

The rut depth is reported as nearside rut depth and offside rut depth averaged over 10m lengths. A value of maximum rut depth is also calculated, which is the maximum of the nearside or offside rut depths within the reporting length.

Figure 3: Calculating rutting from transverse profile

Longitudinal profile – Moving Average Longitudinal Profile Variance (MALPV) and Enhanced Longitudinal Profile Variance (ELPV)

Longitudinal Profile Variance (LPV) is a measure of the ride quality of individual lengths of the road that can be used to assist in the assessment of the likely user opinion of that length of road. SCANNER reports two measures of profile variance – Moving Average (MALPV) and Enhanced (ELPV). TRACS only reports ELPV. Each is reported as the average over each 10m length.

The basis of profile variance is the identification of differences between the measured longitudinal profile and a smoothed version of this profile. This smooth profile is obtained by applying a filter to the measured raw profile over three different lengths: 3m, 10m and 30m to obtain the three measures of profile variance. Both Moving Average LPV and Enhanced LPV apply a similar approach. However, the filters used in ELPV differ from those used in MALPV, so that ELPV is less affected by external influences such as extremes of gradient. ELPV was introduced initially in TRACS as a replacement for MALPV. However, in order to provide continuity in the transfer from one measure to the other, SCANNER currently reports both measures, and the SCANNER RCI is based on MALPV.

Research has shown that Longitudinal Profile Variance can be related to the user’s opinion of ride quality. 3m Longitudinal Profile Variance provides a relatively simple yet effective method of highlighting lengths of the network which users consider to have poor ride quality. 10m Longitudinal Profile Variance also shows reasonable agreement with users, particularly on high-speed roads, and this measure is more likely to reflect the opinions of users of longer vehicles such as buses and trucks. Although 30m Longitudinal Profile Variance is not well related to user opinion, it can be used to show the presence of long undulations associated with defects such as subsidence, and is still reported by TRACS on trunk roads, although it is no longer required by the SCANNER specification on local roads.

Texture in a single line (SMTD)

The surface Texture Depth measured by TRACS and SCANNER accredited survey vehicles is the coarser element of macrotexture and the finer element of mega texture formed by aggregate particles in the asphalt surfacing or by the brushing or grooving of concrete surfaces. Texture Depth contributes to skidding resistance, primarily at medium and high speeds, in 2 ways. Firstly, it provides drainage paths to allow water to be removed rapidly from the tyre or road interface. Secondly, the projections, which contribute to hysteresis losses in the tyre, are an important factor in the braking process.

In TRACS and SCANNER, the texture is measured in the nearside wheel-track using a laser sensor that reports at approximately 1mm longitudinal intervals. The parameter is reported as the average SMTD (Sensor Measured Texture Depth) within each 10m length, which is effectively the RMS height of the texture profile after the removal of wavelengths longer than approximately 300mm.

Cracking - Area of Cracking

SCANNER and TRACS survey vehicles use downward-facing cameras to continuously collect images of the road surface. The video images are processed through automatic, crack-detection software. The cracking is reported as a “crack map” which describes the transverse and longitudinal position of each crack, the length of each crack and the angle of the crack relative to the direction of travel.

The measurement of cracking typically covers a survey width of 3.2m, although this can vary with the survey vehicle (the minimum requirement is 2.9m). The area of cracking is obtained by overlaying the crack map with a 200mm grid covering the whole survey width and counting the number of grid squares containing cracks. The area of cracking over each 10m length is reported as a percentage measure.

The automatic detection of cracking is a demanding task for computer processing systems, and the software employed is subject to a process of continuous improvement to help improve confidence in the cracking data. However, the measurement of cracking is still subject to a degree of variability.

Figure 4: An example of cracking

Further defects measured by TRACS and SCANNER

Transverse Profile – Cleaned Rut Depth (SCANNER only)

SCANNER reports an additional measurement called Cleaned Rut Depths, in each 10m length. Sometimes the measurement of rut depth can be affected by features at the edge of narrower roads such as verges that result in erroneously high rut depths. For Cleaned Rut Depths, an algorithm is applied to identify the road edge, hence removing edge features from the rut calculation. The Cleaned Rut Depth method is still being evaluated in practice to confirm its reliability and robustness. To provide continuity in the transfer from one measure to the other, SCANNER currently reports both measures, and the SCANNER RCI is based on the original rut depth measure.

Transverse Profile – Transverse Unevenness (SCANNER only)

Transverse profile unevenness assesses how much the slope of the transverse profile changes from point to point across the carriageway (Figure 5), calculated as the absolute deviation of the first derivative of the transverse profile and reported each 10m. An even surface would have a low value of unevenness. This measure may be of use when assessing a narrow road where the transverse profile is uneven, but there are no distinct ruts.

Figure 5: Example of Transverse Unevenness

Edge Deterioration (SCANNER only)

SCANNER measurements of edge condition are derived from the measured transverse profile. Algorithms are applied to identify the edge region at the nearside of the traffic lane (Figure 6). Once identified, further algorithms are applied to determine the severity of any stepping at the road edge, and the roughness of the road in the regions close to the road edge. These are reported every 10m. The edge deterioration parameters are not included in the core parameters or the current SCANNER indicator (RCI). However, an edge condition indicator has been included within UKPMS systems to provide an overall report of SCANNER edge condition (see Watson and Wright, 2006).

Figure 6: Identifying the edge region for the assessment of edge deterioration

Longitudinal profile – Bump (TRACS and SCANNER)

User trials have found that users are sensitive to short length “bump” features that cause jolt-like sensations within the vehicle but these are not effectively identified using LPV. The “Bump Measure” has been designed to identify lengths of pavement containing short defects (“bumps”), such as sunken ironwork, potholes, and poor joints (see Figure 7). Such defects can cause jolting discomfort to the road user.

Figure 7: Example of ‘bump’ feature

The Bump Measure is calculated from the Longitudinal Profile and reported over 10m lengths. It reports a simple “Yes” or “No”, with a “Yes” indicating that the length contains one or more bumps. The bump data is intended to complement other measures, for example to assist engineers in prioritising lengths for maintenance.

Texture – in multiple lines (SCANNER only)

Research has shown that reporting texture along multiple measurement lines (across the traffic lane) could be of assistance in identifying lengths of surface deterioration because the variability of texture depth along and across road surfaces can be associated with road surface wear (an “as new” road would be expected to have quite an even texture) - Figure 8. The SCANNER survey therefore reports texture in multiple measurement lines over a minimum of 3 (maximum 40) lines. The measure is reported as the RMST (Root Mean Square Texture), which is the RMS height of the texture profile after the removal of wavelengths longer than approximately 100mm. Various RMST parameters are reported every 10m, covering the regions over, and between, the wheel paths. The texture variability is quite a new measure in SCANNER that aims to assist engineers in identifying lengths of surface deterioration, particularly on lower classes of road.

Figure 8: Variation in average texture depth across the road surface

Texture – Fretting (TRACS only)

The TRACS texture profile data is used to calculate an estimation of the intensity of fretting present on Hot Rolled Asphalt surfaces every 10m. Because the measure applies only to Hot Rolled Asphalt surfaces, the surface type must be known. Furthermore, texture profile data is only measured along a single line in the nearside wheelpath. This measure only estimates the presence of fretting that extends into the wheelpath. Work is ongoing by National Highways to further develop the capability of TRACS in the automated measurement of fretting.

Noise (TRACS only)

The TRACS noise measure applies an algorithm to the texture profile to estimate the level of noise that would be reported at the tyre-road interface within each 10m, which is the noise typically, measured using a close proximity noise measurement system. Whilst the TRACS noise measure can be used to aid the investigation of noise levels, the predicted noise value does not relate directly with the roadside noise level, which is influenced by the volume, type and speed of traffic, and surrounding geometry; nor does it indicate the noise experienced at nearby properties, which is influenced by their distance from the road and the presence of natural or man-made barriers. However, the measure can be used to aid investigation of the noise levels generated by the road/tyre interaction on the pavement. For example, it can be used to provide an indication of the potential benefit of providing lower noise surfacing for sites where noise levels have been identified as an issue.

Cracking – Wheeltrack cracking (both TRACS and SCANNER)

In addition to reporting the area of cracking present over the whole traffic lane, SCANNER and TRACS estimate the extent of cracking present within the region covered by the wheel paths. The approach taken by each system differs. SCANNER reports the total length of the wheeltrack affected by wheel track cracking (as a percentage of each 10m length). TRACS reports the total area of wheeltrack cracking present in each 10m length (as a percentage of the total area of the wheeltrack). Because the measure of wheeltrack cracking is effectively a sub-set of the total area of cracking present, wheeltrack cracking is not currently included in the SCANNER RCI.

Geometry (both TRACS and SCANNER)

The 3 measures of road geometry are delivered every 10m, as Gradient, Crossfall and Radius of Curvature. Gradient and Crossfall are reported as percentage values, whilst Radius of Curvature is reported in metres.

Retro-reflectivity (TRACS only)

The TRACS survey measures the retro-reflectivity of longitudinal and in-lane road markings. A device fitted to the vehicle directs light at the road markings and reports the amount of light reflected, which provides an indication of the night-time visibility of the marking every 10m. The data is used to support the assessment of the condition of road markings and the identification of lengths for investigation.

Assessing road condition using TRACS and SCANNER data

TRACS

The condition of trunk roads is assessed by comparing the values measured by TRACS for each length (100m) against thresholds defined in the England National Application Annex to volume CS 230 of the Design Manual for Roads and Bridges. Four condition categories are used in the interpretation of TRACS condition data. The categories are used to grade pavement condition starting at Category 1, which indicates no visible deterioration, up to Category 4, which indicates severe deterioration. When using TRACS data highways engineers are advised that any length for which the defect value falls above Category 3 contains moderate deterioration and is at a warning level of condition where the need for maintenance must be investigated. If the level is at Category 4, detailed investigations should be carried out on the deteriorated lengths at the earliest opportunity.

Surface Condition of Trunk Roads

The data on the surface condition of trunk roads provided in the publication describes the percentage of the trunk road network that may be in need of maintenance. It is calculated using both TRACS data, and skid resistance data provided by SCRIM surveys (see section 4 for further information on the measurement skid resistance).

The data on the surface condition of trunk roads is obtained using a similar approach to that described above, where the condition is assessed by comparing the values measured for each length (100m) against thresholds. To estimate the percentage of the trunk road network that may be in need of maintenance it is assumed that not all lengths with condition exceeding the Category 3 thresholds defined in the DMRB will be in need of maintenance. However, all lengths exceeding the Category 4 thresholds should be considered to be in need of maintenance. Therefore, an intermediate set of thresholds was established in 2004 for the assessment of TRACS data to indicate the levels of condition more likely to require maintenance in the short term, as shown in Figure 9.

Figure 9: TRACS surface condition thresholds

Defect	Threshold
Rutting	15.5mm
Texture	0.6mm
Enhanced Longitudinal Profile Variance (LPV) – 3m  Motorways and Rural Dual Carriageways Urban Dual Carriageways Rural Single Carriageway Roads Urban Single Carriageway Roads	3.3mm squared 3.85mm squared 3.85mm squared 6.55mm squared
Enhanced Longitudinal Profile Variance (LPV) – 10m Motorways and Rural Dual Carriageways Urban Dual Carriageways Rural Single Carriageway Roads Urban Single Carriageway Roads	10.6mm squared 15.7mm squared 15.7mm squared 27.45mm squared
Enhanced Longitudinal Profile Variance (LPV) – 30m  Motorways and Rural Dual Carriageways Urban Dual Carriageways Rural Single Carriageway Roads Urban Single Carriageway Roads	88.0mm squared 98.0mm squared 98.0mm squared 145.0mm squared

For the assessment of SCRIM data an equivalent Category 3a value was determined based on advice given in section HD28/04 of the Design Manual for Roads and Bridges. The Category 3a threshold is generally equal to the (Investigatory Level – 0.05), see section 4.

In the calculation, if one or more of the condition parameters exceeds any of the thresholds, the length is flagged (note that for texture the length is flagged if the texture parameter falls below the threshold, because a decrease in texture represents a deterioration in condition). The value reported in the publication has been calculated by summing the length of pavement that triggers any of the specified threshold levels and expressing the sum as a percentage of the network length.

SCANNER

Individual SCANNER parameters can be used to identify lengths containing particular types of defect. However, the SCANNER RCI (Road Condition Index) was developed through the SCANNER research programme to combine SCANNER defects into a single value to assist in the assessment of road condition. The approach used to combine the SCANNER defects was developed by Cartwright and Pickett (2004). This was used with an initial set of thresholds and weightings to calculate the ‘original RCI’ for 2006 and 2007. Further research (McRobbie, Walter, Read, Viner and Wright, 2007) led to new thresholds giving a ‘revised RCI’ which has been used since 2008 and is described herein.

The revised SCANNER RCI is calculated using a sub-set of the parameters measured by SCANNER. In summary, these are:

• maximum rut depth

• 3m Moving Average Longitudinal Profile Variance

• 10m Moving Average Longitudinal Profile Variance

• whole carriageway cracking

• texture depth

To obtain an RCI value each parameter is scored between 2 thresholds – a lower threshold below which there is no need to consider maintenance, and an upper threshold above which further deterioration does not increase the score. These thresholds were based on engineers’ experience of each parameter. The score increases linearly between the lower and upper threshold from zero at the lower threshold to 100 at the higher. Figure 10 demonstrates the application of this method for rut depth.

The score for each parameter is then multiplied by 2 factors, each having a value between 0 and 1. One factor reflects the “relevance” or importance of the measurement to the maintenance condition of the road. The other reflects the “reliability” of the method of measurement. The result is a weighted score for each parameter for the 10m subsection.

Note that, to avoid Longitudinal Profile Variance having a disproportionate affect on the reported condition, the weighted scores for 3m Moving Average Longitudinal Profile Variance and 10m Moving Average Longitudinal Profile Variance are compared, and only the largest of these 2 scores is taken forward to contribute to the calculation of the RCI. The same is true for rutting, with only the maximum of the offside and nearside rut depths being taken forward to the calculation of the RCI.

The weighted scores are summed to give a single RCI value for each 10m subsection length, representing the overall condition. The SCANNER RCI values reported by SCANNER can be used by highway engineers (often by displaying the data on a map background in a GIS) to identify lengths of the network in need of further, more detailed, investigation.

Figure 10: Scoring rutting for the SCANNER RCI

Example RCI calculation

Below is a summary table of an example RCI calculation. Figure 10 shows how the two rutting variables (LRRT and LLRT) have been scaled to a RCI value between 0 and 100. For instance, a recorded value of 11.5mm is 15% of the way between 10mm and 20mm so this produces an RCI value of 15. The other variables have been scaled using a different linear interpolation curve. For this particular data point, there was no cracking element recorded and this goes forward as a zero in the RCI calculation. LV10 and LV3 are the two longitudinal profile variance variables, LTRC is cracking and LLTX is texture depth.

Variable	Recorded Value	RCI Value	Reliability Weighting	Importance Weighting
LRRT	11.5	15	1	1
LLRT	13.4	34	1	1
LLTX	0.29	100 (the actual figure is 102.5, but this is scaled back to a maximum 100)	1	0.75
LV10	41.52	58.63	1	0.6
LV3	4.79	13.17	1	0.8
LTRC	Null	-	0.6	1

The RCI is calculated by summing the weighted individual RCI values, taking the maximum of the weighted LRRT or LLRT and the maximum of the weighted LV10 or LV3 variables forward as demonstrated below:

RCI = max((LRRT * 1* 1) , (LLRT * 1 * 1)) + (LLTX * 1 * 0.75) + max((LV10 * 1 * 0.6) , (LV3 * 1 * 0.8)) + (LTRC * 0.6 * 1)

In this case, this becomes:

RCI = (34 * 1 * 1) + (100 * 1 * 0.75) + (58.63 * 1 * 0.6) + (0 * 0.6 * 1)

       = 34 + 75 + 35.18 + 0

       = 144.18

A large RCI score such as this indicates this section of the road is in ‘poor’ condition and is likely to require maintenance soon.

SCANNER National Indicators for local roads

The RCI values for each 10m length can be summed to determine the overall percentage of the 10m lengths within the network falling into 3 categories:

• “GREEN” - lengths where the carriageway is generally in a good state of repair (low RCI values). Green lengths have an RCI score below 40

• “AMBER” - lengths where some deterioration is apparent which should be investigated to determine the optimum time for planned maintenance treatment (mid-range RCI values). Amber lengths have an RCI score over 40 and below 100

• “RED” - lengths in poor overall condition which are likely to require planned maintenance soon (for example, within a year or so) on a “worst first” basis (high RCI values). Red lengths have an RCI score of 100 or over

Figure 11: Examples of roads which would fall under the green (left), amber (centre) and red (right) categories

In England, local authorities can use UKPMS to obtain the proportion of the network in the “red” category, which is reported as the Single Data List items 130-01 (‘A’ class roads) and 130-02 (other classified roads) for their network. These were formerly National Indicators 168 and 169 respectively.

The parameters, thresholds and weightings that define the calculation of the RCI are published on the DfT website as a “Weighting Set” (see below).

Further information on TRACS and SCANNER

TRACS

Design Manual for Roads and Bridges (DMRB) – HD 29/08, Volume 7, Section 3, Part 2, “Traffic speed condition surveys”, National Highways.

Design Manual for Roads and Bridges (DMRB) - CS230 “Pavement maintenance assessment procedure”, National Highways.

SCANNER

SCANNER user guide and specification, Volume 1, Introduction to SCANNER surveys. (Roads Board).

SCANNER user guide and specification, volume 2, Procuring SCANNER Surveys. (Roads Board).

SCANNER user guide and specification, volume 3, Using SCANNER Survey Results. (Roads Board).

SCANNER user guide and specification, volume 4, Technical Requirements for SCANNER Survey Data and Quality Assurance. (Roads Board).

SCANNER user guide and specification, volume 5, Technical Requirements for SCANNER Survey Parameters and Accreditation. (Roads Board).

All above found on UKRLG SCANNER advice.

E. Benbow, K. Nesnas and A.Wright (2006), Shape (surface form) of local roads. TRL published project report PPR131. 

Reporting condition and indicators

Cartwright, R. A and Pickett, A. (2004). TTS Defects Index Preliminary Analysis – Final Report version 2. (Chris Britton Consultancy). 

Watson, P and Wright, A. a. (2006). Edge deterioration on Local Roads. TRL Published Project Report PPR084. TRL Limited.

McRobbie, S., Walter, L., Read, C., Viner, H and Wright, A. (2007). Developing SCANNER Road Condition Indicator parameter thresholds and weightings. TRL Published Project Report PPR199. (TRL Limited, Wokingham).

Design Manual for Roads and Bridges (DMRB) – CS228 “Skidding Resistance”.

UKPMS Technical Note 36 and 37 – Calculation of RCI values for BVPI and National Indicators.

UKPMS Weight Sets and RCI Description.

UKPMS Document No 078 (2005), SCANNER Road Condition Indicator, Description of Processing aimed at Users.

UKPMS Document No 070 (2006), SCANNER Road Condition Indicator, Implementation Guidelines.

Contact details

Road condition statistics

Email [email protected]

Media enquiries 0300 7777 878