How roughness survey data gets into RAMS: profiler run to register

Roughness survey data gets into a road asset management system (RAMS) through three hand-offs between South African road manuals: TMH 13 (Committee Draft Final, May 2016) for the profiler run and its 100 m averages (COTO, 2016a), TMH 18 (Draft Version 4, October 2014) for the combined instrument data (.cid) file (COTO, 2014), and TMH 22 (Draft, March 2013) for the register (COTO, 2013). The value that travels is the International Roughness Index (IRI), in metres per kilometre (m/km); the .cid file carries it with road, lane, kilometre window and date (COTO, 2014).

Key facts

• Section D.4.3 of Draft TMH 22 (March 2013), a South African manual, notes that profilers can record International Roughness Index (IRI) every 10 m and recommends 100 m averaging.
• TMH 13 Part A (Committee Draft Final, May 2016), a South African manual, advises against interpreting International Roughness Index (IRI) over lengths under 30 m and recommends 100 m averages.
• TMH 18 Section 5.6 puts roughness, rutting and texture in one combined instrument data (.cid) record only when they share road, lane code, start and end kilometre and measurement date.
• TMH 18 Committee Draft 5.0 adds an authority identifier as the first .cid column and renames headings used in Draft Version 4, and neither version requires a file header.
• Draft TMH 22 (March 2013) Section B.1.4 states that the direction of the line representing a road must match the direction of its kilometre distances.
• Table D-8 of Draft TMH 22 (March 2013) lists roughness storage per 100 m, and its note asks for raw 10 m values to be kept as supplied.

What is the route from a profiler run to a register value?

The route has three hand-offs: TMH 22 hands the measurement to TMH 13 (COTO, 2013), TMH 13 Part C hands the report format to TMH 18 (COTO, 2016a), and the TMH 18 file loads into the TMH 22 register (COTO, 2014) (COTO, 2013). TMH (Technical Methods for Highways) is the series of road manuals issued under the Committee of Transport Officials (COTO). TMH 13 (Automated Road Condition Assessments) guides the planning, execution and control of the measurement, TMH 18 (Road Asset Data Electronic Exchange Formats) sets out the file that carries the result, and TMH 22 (Road Asset Management Manual) describes the RAMS that stores it.

A high-speed profiler is a vehicle-mounted device that measures the road profile and calculates IRI for each wheelpath, the line one wheel follows (COTO, 2016a) (COTO, 2016d). In this guide, the register is the RAMS database: the Part B road inventory of TMH 22 with Part D condition data stored against it (COTO, 2013), not the fixed asset register used for valuation.

The roughness value changes scale once, when 10 m readings become a 100 m average; road, lane, kilometre window, direction and date must stay identical at every hand-off.

Who fixes the averaging window and the location rules before the run?

The survey specification fixes the averaging window, and the register supplies the location rules.

TMH 22 Section D.4.3 observes that modern profilers can record IRI at 10 m intervals and recommends averaging over 100 m sections. Table D-8 lists the roughness procedure: specified survey direction, slow lane, both wheelpaths, 10 m test interval, storage per 100 m, date, and direction of measure reported (COTO, 2013).

TMH 13 Part A says the specification should define the averaging procedure and the segment length for reporting, with fields such as section name, kilometre limits, direction and survey date (COTO, 2016d). See how to specify a road condition survey.

For location, TMH 22 Section B.1.3 notes that rural roads are often cut into sections that each restart at a zero kilometre, and Section B.1.4 adds that the digital line for a road must run in the same direction as its kilometre distances, with the surveyed length kept beside any differing official kilometre positions (COTO, 2013). These give the processor a datum: the chainage, or distance along the road, at which the first 10 m reading is placed.

Narrated film · 1:24

Profiler run to register value

A synthetic 300 m stretch: thirty 10 m readings collapse into three 100 m records as the value passes from TMH 13 to TMH 18 to TMH 22.

Read the film transcript

How does a profiler's roughness reading become a value in the road asset register? Three South African road manuals carry the International Roughness Index, the IRI: TMH 13 guides measurement, TMH 18 sets out the file, TMH 22 describes the register. At each hand-off the value must keep its road, lane, kilometre window and date. A profiler reads this synthetic stretch every ten metres, from four point two to four point five kilometres. That is thirty readings in three windows, with one high reading in the third. TMH 22 recommends one hundred metre averages, here one point nine, three point one and two point eight metres per kilometre. The third average no longer shows where the high reading sits, so keep the ten metre values. In TMH 18 Draft Version 4, each window is one record, and IRI, rut and texture share road, lane, kilometre window and date. Committee Draft five point zero adds an authority column, so a Draft Version 4 file read against it shifts every field. The survey starts at four point two kilometres, but the distance meter reads zero. Without that offset, the first window is written at zero, still loads, and sits four point two kilometres from where it was measured. Only the averaging changes the value's scale; its place on the road must never change. Next, TMH 22 Part E turns stored values into indices, and TRH 12 describes pavement rehabilitation investigation.

How does a 10 m reading become a 100 m value?

A processor averages the ten 10 m readings in each 100 m window and stores one record per window (COTO, 2013).

The IRI algorithm filters out wavelengths outside the 1.3 m to 30 m band, so TMH 13 Part A advises against interpreting IRI over lengths shorter than 30 m and recommends 100 m averages (COTO, 2016d). The window recurs for rut depth, which TMH 13 Part D stores at 10 m and reports at 100 m (COTO, 2016e).

Synthetic example. Every value is invented. Road SYN-S1, lane P1, was profiled on 14 September 2026 over register kilometres 4.200 to 4.500, recording left-wheelpath IRI every 10 m, thirty values in m/km:

• Window 1, 4.200 to 4.300 km: 1.8, 2.0, 1.9, 2.1, 1.7, 2.0, 1.9, 2.2, 1.8, 1.6. Sum 19.0, mean 1.9.
• Window 2, 4.300 to 4.400 km: 3.0, 3.2, 3.1, 3.3, 3.0, 3.1, 3.2, 2.9, 3.1, 3.1. Sum 31.0, mean 3.1.
• Window 3, 4.400 to 4.500 km: 2.4, 2.6, 2.5, 2.7, 5.0, 2.8, 2.5, 2.6, 2.4, 2.5. Sum 28.0, mean 2.8.

Each mean is the sum divided by ten, because the readings cover equal lengths. The 5.0 m/km reading at 4.440 to 4.450 km lifts window 3 by about 0.2 m/km, and the averaged record no longer shows where. The note under TMH 22 Table D-8 asks for the 10 m values to be stored as supplied (COTO, 2013), although TMH 13 Part A advises against interpreting any 10 m value alone (COTO, 2016d).

Each window becomes one record. The table shows five Draft Version 4 headings from TMH 18 Section 5.6, with invented rut depth and texture (mean profile depth, MPD) values for the same windows. All three share road SYN-S1, lane code P1 and the date.

START_KM	END_KM	AVG_IRI_L (m/km)	AVG_RUT_L (mm)	AVG_MPD_L (mm)
4.200	4.300	1.9	5.0	0.90
4.300	4.400	3.1	7.0	0.80
4.400	4.500	2.8	6.0	0.90

What does one .cid record carry, and what must it share?

TMH 18 Section 5.6 describes the combined instrument data file as one record type for roughness, rutting and texture. A single-file submission needs all three measured at the same time and location, because they share road, lane code, start and end kilometre and measurement date (COTO, 2014). All three therefore need the same window; see IRI, rut depth and texture depth compared.

TMH 18 Section 5.6 states no window length: the Draft Version 4 example record runs from 3.60 to 3.70 km, the Committee Draft 5.0 example from 0 to 2 km (COTO, 2014) (COTO, 2016i). The 100 m window comes from TMH 22 and the contract.

Neither version requires a file header, because the file definitions are predefined. Committee Draft 5.0 (February 2016) adds an authority identifier as the first column and renames headings, for example TEST_DATE for MEASURE_DATE, so a Draft Version 4 file read against it shifts every field by one (COTO, 2014) (COTO, 2016i). Both versions lack a field for profiler class, validation date or averaging rule; keep them with the delivery.

How do road, lane, kilometre and direction keep the value in the right place?

TMH 18 Section 2.1 makes the road the base reference system: its identifier is unique and related records use its kilometre measure. Section 5 discards values outside the road's start and end kilometre, and Section 2.2.2 maps data to a road version by measurement date (COTO, 2014). Section 4.1 defines the primary direction as the side on which kilometres increase towards north or east, and Section 4.2 numbers lanes from the fast lane. TMH 22 Table D-8 asks for the slow lane (COTO, 2013), whose code depends on the lane layout, so take it from the lane configuration.

TMH 13 Part B Section B.4.3 lists checks on delivered files: section ends, survey direction and reference distances agree with the road definition, and position points are evenly spaced at the reporting interval (COTO, 2016c).

In the synthetic example, SYN-S1 starts at 0.000 km on the register and the distance meter was zeroed at the survey start, 4.200 km. A processor that forgets to add 4.200 km writes the first window as 0.000 to 0.100 km. That lies inside the road, so it loads 4.2 km from where it was measured. The insight on chainage, direction and wheelpath covers a reversed survey.

What are the steps from profiler run to register?

1. Fix the window and the rules in the specification. Name the 100 m window, averaging procedure, lane, wheelpaths, direction and date, as TMH 22 Table D-8 and TMH 13 Section A.4.3.7 list them (COTO, 2013) (COTO, 2016d).
2. Take the road, kilometre limits and direction from the register. Copy them from the TMH 22 Part B inventory.
3. Validate the profiler and the positioning. Use validation sections (TMH 13 Parts B and C); if a control test fails, discard and remeasure the data since the last passed test (COTO, 2016a) (COTO, 2016c).
4. Run the profiler and record 10 m values. Drive the specified lane and direction, recording IRI, rut depth and MPD with position and operator flags.
5. Check start, end, direction and distance. Compare section ends, survey direction and reference distances with the road definition (COTO, 2016c).
6. Average to 100 m windows and keep the 10 m values. Average every measure over the same windows and store the raw values as supplied.
7. Write one record per window and name the TMH 18 version. Share road, lane, kilometre window and date across IRI, rut depth and MPD; see how to check a TMH 18 file set before sending.
8. Load against the right road version and reconcile. Load by measurement date against the road version and its kilometre range (TMH 18 Section 2.2.2) (COTO, 2014), then check record counts, duplicates and key fields (TMH 22 Section D.5.12) (COTO, 2013).

What do the manuals not decide, and what is commonly misread?

The manuals leave three decisions open, and three misreadings recur.

What the manuals leave to the specification

• Averaging formula and partial windows. TMH 13 Part A leaves the averaging procedure to the specification (COTO, 2016d); the passages cited here do not say how to treat a window with missing 10 m readings.
• One value per lane. The .cid record keeps left and right wheelpath IRI apart. The COTO Standard Specifications, Chapter 20 (Draft Standard, October 2020), average them for judging constructed work (COTO, 2020a), so both forms circulate.
• What the value means for treatment. A stored IRI shows where to look and does not select a treatment; see the insight on interpreting IRI in South Africa.

Common misreadings

• "TMH 18 fixes the 100 m window." Section 5.6 states no length.
• "MPD is defined over 100 m." MPD is the mean profile depth over a 100 mm baseline (COTO, 2016f); the 100 m window only averages those values.
• "Look for the formats in TMH 20." TMH 22 and TMH 13 Parts B and D refer format questions to TMH 20 (COTO, 2013) (COTO, 2016c) (COTO, 2016e), which TMH 22 Section A.2 titles "Standard for Road Network Data", the running head of TMH 18 Draft Version 4 (COTO, 2014); Part C names TMH 18 (COTO, 2016a).

What evidence should each stage leave behind?

• Specification: the clause naming window, averaging, lane, wheelpaths, direction and date.
• Validation: validation reports, control-test results and the positioning check log.
• Raw data and processing: the 10 m values as supplied, operator flags, the averaging procedure and the window table.
• Exchange: the .cid file, its TMH 18 version and the definition sheet for any codes (COTO, 2016a).
• Register: the load report with record counts, road version and kilometre range.

Which manual governs each stage

Stages of the journey, the manual that governs each one, and the evidence each stage should leave behind.

Stage	Governing manual	Evidence this stage produces
1. Specify the survey	TMH 22 Part D	The specification clause that names the 100 m window, averaging procedure, lane, wheelpaths, direction and date.
2. Plan and validate the survey	TMH 13 Part A	The survey plan, equipment specification and validation requirements, with the validation reports for the profiler.
3. Fix position and direction	TMH 13 Part B	The positioning check log: section starts and ends, survey direction, reference distances and point spacing against the road definition.
4. Measure and average	TMH 13 Part C	The 10 m IRI values as supplied, the operator flags, and the 100 m averages with their kilometre limits.
5. Exchange the file	TMH 18	The .cid file with its TMH 18 version named, one record per window sharing road, lane, kilometre window and date.
6. Register the record	TMH 22 Part B	The load report showing road version, kilometre range, record counts and key-field checks.

The journey, step by step

1. Fix the window and the rules in the specification

   Name the 100 m reporting window, the averaging procedure, the lane, both wheelpaths, the survey direction and the date in the contract. TMH 22 Table D-8 and TMH 13 Part A Section A.4.3.7 list what to define.

2. Take the road, kilometre limits and direction from the register

   Copy the road identifier, the start and end kilometre and the kilometre direction from the TMH 22 Part B inventory before the run, so the survey and the register share one datum.

3. Validate the profiler and the positioning

   Check the profiler on validation sections and the positioning equipment before the survey, as TMH 13 Parts C and B describe. If a control test fails, discard and remeasure the data collected since the last passed test.

4. Run the profiler and record 10 m values

   Drive the specified lane and direction and record IRI, rut depth and mean profile depth every 10 m, with position and operator flags for unusual conditions.

5. Check start, end, direction and distance

   Compare section starts and ends, survey direction and reference distances with the road definition, and look for gaps or jumps in position, before any processing.

6. Average to 100 m windows and keep the 10 m values

   Average every measure over the same 100 m windows using the procedure in the specification, and store the raw 10 m values as supplied.

7. Write one record per window and name the TMH 18 version

   Give IRI, rut depth and texture the same road, lane code, start and end kilometre and measurement date in one .cid record, and state whether the file follows Draft Version 4 or Committee Draft 5.0.

8. Load against the right road version and reconcile

   Load by measurement date against the matching road version and its kilometre range, then check record counts, duplicate records and key fields against what was delivered.

Frequently asked

How do 10 m roughness readings become the 100 m IRI value in a road asset register?

A processor averages the ten 10 m International Roughness Index (IRI) readings in each 100 m window, as the South African manual Draft TMH 22 (March 2013) Section D.4.3 recommends, and stores one record per window against the road, lane and kilometre. The specification should state the averaging procedure, and the raw 10 m values should be kept.

Does TMH 18 say how long the averaging window must be?

TMH 18 (Road Asset Data Electronic Exchange Formats) states no window length. Section 5.6 describes the combined instrument data (.cid) record without one, the example records in Draft Version 4 and Committee Draft 5.0 differ in length, and the standard assessment lengths in Section 5.1 do not mention the .cid file. The 100 m reporting interval comes from TMH 22 Part D and the survey contract, so state it in the delivery note.

What must IRI, rut depth and texture share in one .cid record?

In a single combined instrument data (.cid) file under TMH 18 Section 5.6, roughness (International Roughness Index, IRI), rutting and texture (mean profile depth) share the road identifier, lane code, start and end kilometre and measurement date. If they were measured at a different time or place, they do not belong in the same record.

Which TMH 18 version should a roughness file follow?

TMH 18 does not choose a version for the receiver, so agree the version with the receiving road authority and name it in the delivery note. TMH 18 exists as Draft Version 4 (October 2014) and Committee Draft 5.0 (February 2016). Committee Draft 5.0 adds an authority identifier as the first column, and neither version requires a file header, so the file itself does not show which order it follows.

Why does the kilometre datum matter when loading roughness data?

The kilometre value places a roughness record on the road, so a wrong offset that stays inside the range still loads, in the wrong place. Draft TMH 22 Section B.1.4 says the line direction must match the kilometre direction, TMH 13 Part B lists checks on section ends and reference distances, and TMH 18 discards values outside the road's start and end kilometre.

How do TMH 13, TMH 18 and TMH 22 connect for roughness data?

TMH 13 guides the planning, execution and control of the measurement, TMH 18 sets out the combined instrument data (.cid) file that carries the 100 m values to the receiver, and TMH 22 describes the register that stores them. The roughness value changes scale once, from 10 m readings to a 100 m average, while road, lane, kilometre window, direction and date stay identical at every hand-off.

Where this connects

The next manual is TMH 22 Part E, which turns stored IRI values into indices; Section E.3.4 expresses the riding quality of a road link from IRI values per 100 m segment (COTO, 2013). After it comes TRH 12 (Technical Recommendations for Highways, Draft 1997), which covers the project-level investigation and design of flexible pavement rehabilitation that a network result may prompt (Department of Transport, 1997).

Last reviewed on 21 October 2026 against TMH 13, TMH 18, TMH 22 and the COTO Standard Specifications, in the editions named above.

References

1. COTO (2016a). TMH13: Automated Road Condition Assessments. Part C: Roughness. Committee Draft Final, May 2016. Original source.Sections C.2–C.5; Table C.5 is for calibration-section selection
2. COTO (2016c). TMH13: Automated Road Condition Assessments. Part B: Positioning. Committee Draft Final, May 2016. Original source.Positioning and distance measurement guidance
3. COTO (2016d). TMH 13: Automated Road Condition Assessments. Part A: General. Committee Draft Final, May 2016. Original source.Sections, tables and clauses are cited inline in each guide
4. COTO (2016e). TMH 13: Automated Road Condition Assessments. Part D: Rutting. Committee Draft Final, May 2016. Original source.Sections, tables and clauses are cited inline in each guide
5. COTO (2016f). TMH 13: Automated Road Condition Assessments. Part E: Skid Resistance and Texture. Committee Draft Final, May 2016. Original source.Sections, tables and clauses are cited inline in each guide
6. COTO (2014). TMH 18: Road Asset Data Electronic Exchange Formats. Draft Version 4, October 2014. Original source.Sections, tables and clauses are cited inline in each guide
7. COTO (2016i). TMH 18: Road Asset Data Electronic Exchange Formats. Committee Draft 5.0, February 2016. Original source.Sections, tables and clauses are cited inline in each guide
8. COTO (2013). Draft TMH 22: Road Asset Management Manual. Draft, March 2013. Original source.Sections, tables and clauses are cited inline in each guide
9. COTO (2020a). Standard Specifications for Road and Bridge Works for South African Road Authorities. Draft Standard, October 2020. Original source.Sections, tables and clauses are cited inline in each guide
10. Department of Transport (1997). TRH12: Flexible Pavement Rehabilitation Investigation and Design. Draft TRH12, Pretoria: Department of Transport for COLTO. Original source.Chapter 1: rehabilitation design process; investigation, evaluation, analysis and economic appraisal

Use the manual edition and project requirements that apply to your work. Worked examples use synthetic values and are not site-specific design advice.