How to calculate a visual condition index (VCI) from TMH 9 ratings

A visual condition index (VCI) for a road pavement is a single number from 0 to 100 that condenses the degree and extent ratings of many distresses on one assessment segment. The South African road manual TMH 9 (Technical Methods for Highways, Committee Draft Final, May 2016) supplies those ratings but no formula, so a calculation should name the published variant it follows and use that variant's weights. (COTO, 2016b)

Key facts

• TMH 9 (Committee Draft Final, May 2016) records degree and extent ratings for each distress on an assessment segment and contains no visual condition index formula, weight or transform.
• A visual condition index (VCI) for a road pavement follows a named variant, such as Draft TMH 22 Appendix J-6 (March 2013), and that variant's own weight table.
• TMH 22 Appendix J-6 continues the Draft TRH 22 section 6.2.2 form, and the modified Technical Addendum A.3 method in Draft TRH 22 was not carried into TMH 22.
• The same TMH 9 ratings give different visual condition index (VCI) values under different variants, so each VCI should be quoted with its variant, weight table and ratings edition.
• An item rated absent stays in the denominator of a visual condition index (VCI), while an item that was not rated follows the named variant's own rule.
• TMH 18 flexible-pavement visual condition files (Draft Version 4, 2014; Committee Draft 5.0, 2016) hold the VCI to one decimal place and have no field naming the variant.

What is a visual condition index for a road pavement?

A visual condition index for a road pavement summarises the ratings of one segment so that segments can be compared, ranked and followed over time. TMH 9 rates one segment at a time and recommends that the authority predefine each segment's start and end distance along the road (chainage). For each distress (a visible sign of pavement deterioration, such as cracking, rutting or potholes) the assessor records a degree (how severe it is, rated 0 to 5) and an extent (how widespread it is, rated in five classes). See degree and extent as evidence. (COTO, 2016b)

Draft TRH 22 (Technical Recommendations for Highways, Pavement Management Systems, 1994) says the VCI takes no account of traffic or functional classification. (CSRA, 1994) Draft TMH 22 (Road Asset Management Manual, March 2013) calls the VCI an arithmetic aggregation of distress ratings determined in accordance with TMH 9, a good indicator of trends that inherits the weaknesses of averaging. (COTO, 2013)

What does TMH 9 give you, and where does it stop?

TMH 9 supplies the degree and extent ratings and stops there: none of its Parts A to E sets out a visual condition index formula, weight or transform, although Part A names visual condition indices as a main output and its annexure quotes one province's deduct values for checking assessors. Part A sets out the attributes of distress, the standard segments, assessor accreditation and quality management; Part B, for flexible pavements, describes each distress and carries the assessment form. (COTO, 2016b) (COTO, 2016j)

Part B notes that macro texture and voids are not used in the condition index calculation and treats the overall pavement condition rating as a verification check, not a processing input. The arithmetic sits in TMH 22 Appendix J-6 and Draft TRH 22 section 6.2.2. (COTO, 2016j) (COTO, 2013) (CSRA, 1994) Producing the ratings is the subject of how to run a TMH 9 visual assessment.

Narrated film · 1:28

From TMH 9 ratings to a road VCI

A narrated walkthrough of verifying ratings, naming the variant, weighting, normalising, transforming and recording the method, using invented weights.

Read the film transcript

How do ratings from the South African road manual TMH 9 become a visual condition index, the VCI? The VCI condenses degree and extent ratings of many distresses on one segment into one number. TMH 9, a 2016 committee draft, supplies those ratings but no formula, no weights and no transform. Verify the ratings, then name the variant the client's terms of reference specify, such as Draft TMH 22 Appendix J-6. Take each weight from that variant's own table, multiply degree, extent and weight, then normalise against the maximum possible. Apply the variant's transform, if it has one, and write the method beside the number. In this synthetic example we use invented weights, not any manual's, and three distresses contribute forty-eight, thirty-six and sixteen. They total one hundred against a possible five hundred, so the preliminary index is eighty. Invented variant X reports eighty and invented variant Y sixty-four from the same ratings, so neither can be read without its variant name. Never mix one variant's weights with another's transform, and keep an item rated absent in the denominator. Record the variant, weight table and rounding beside every value, since TMH 18 visual condition files have no variant field. Whichever variant you choose, name it in full every time you quote a VCI. Next, see how TMH 22 places the index among condition indices, and how TMH 18 passes it to the authority.

Which variant of the visual condition index should you name?

Name the variant the client's terms of reference choose; TMH 9 does not choose. A variant is a published calculation with its own item list, weight table and sometimes a transform. The table ranks none.

Variant to name	Where it is set out	What distinguishes it
TMH 22 Appendix J-6	Draft TMH 22 (March 2013)	Continues the section 6.2.2 form with a revised item list and its own weight table
Draft TRH 22 section 6.2.2	Draft TRH 22 (1994), equations 6.1 to 6.3, Table 6.1	The proposed formula, weight set and transform; categories follow in section 6.2.3
Draft TRH 22 Technical Addendum A.3	Draft TRH 22 (1994)	Adds an extent weight factor and a small-degree factor; A.1 says the addendum's techniques were not yet accepted practice
ReCAP (Research for Community Access Partnership) guideline Part B, section B.5	Raters' Guideline for low-volume rural roads (Final, September 2020)	Own weight and category tables; no transform
An authority's own weight set	The authority's procedure	Name the weight set and its edition

Draft TRH 22 proposes keeping deviation from its default weight set to a minimum. The ReCAP guideline (not a TMH manual) says each road agency should verify the weightings and modify them to its own assessment of each defect. (CSRA, 1994) (ReCAP, 2020a) The deduct-point indices CISURF (surfacing) and CIPAVE (pavement; also printed CIFPAVE) in TMH 22 Part E and Appendix J-5 are a different family, not VCI variants. (COTO, 2013)

How do you calculate a visual condition index from TMH 9 ratings, step by step?

Work through nine steps per segment; each ends with a record.

1. Confirm which ratings you hold. Check the TMH 9 edition, part, form and segment; flexible pavements follow Part B. Record: edition, part, form, road, segment, lane, assessor, date.
2. Verify the ratings before any arithmetic. TMH 9 Part A (A.5.4) says quality checks can cover the calculated indices and individual degree and extent ratings. Draft TMH 22 (D.5.11, Appendix J-4) gives an example routine that flags conflicting ratings. Return a flagged rating to the form or photographs; never edit a number to clear a flag. Record: checks run, flags, resolutions.
3. Choose the variant and name it. Use the client's named variant; where none is named, choose one from the table, say why and keep it for the whole programme. Quote it in full: "VCI (TMH 22 Appendix J-6, Draft March 2013)". Record: variant, manual, edition, clause.
4. Fix the item list, the unit and the missing-data rule. Use the named variant's item list. Calculate one index per segment and lane as rated (Part A, A.5.1, concentrates a multi-lane assessment on the worst lane). Zero means assessed and absent, never not looked at. Record: item list, lane, missing-data rule.
5. Take the weights from the named table. Copy each weight from that table and any default extent for functional items, such as riding quality, from the variant's text; defaults differ between forms. Record: table identifier and any local change.
6. Weight each rating. Multiply the degree, the extent and the weight for each item, using the variant's own product; the Addendum A.3 form adds two further factors. Keep contributions unrounded. Record: every item's contribution.
7. Normalise against the maximum possible. Add the contributions and divide by the total the item list would give at the maximum ratings the variant defines (Draft TRH 22 and J-6 average a distress's alternative weights). Subtract that ratio from 1 and multiply by 100 for the preliminary index. An item rated absent stays in the denominator; an item not rated follows the variant's rule (ReCAP Annex C drops it). (ReCAP, 2020a) Record: numerator, denominator, preliminary index.
8. Apply the variant's transform, if it has one. The Draft TRH 22 and TMH 22 forms transform the preliminary index to a percentage scale with two printed constants; the ReCAP form reports its normalised value directly. Apply only the named variant's transform; round once, at the end. Record: transformed value, rounding rule.
9. Record the method with the number. Write the variant, weight table, ratings edition and rounding beside the value wherever it is stored or quoted. The flexible-pavement visual condition files of TMH 18 (Road Asset Data Electronic Exchange Formats; Draft Version 4, October 2014; Committee Draft 5.0, February 2016) hold the VCI to one decimal place but have no field naming the variant. (COTO, 2014) (COTO, 2016i) Record: the method line below.

What does a worked example of a visual condition index calculation look like?

Synthetic example. Every value is invented, including the weights, which are not those of TMH 22 Appendix J-6, Draft TRH 22, the ReCAP guideline or any other manual. It takes Synthetic Road S1, one segment from chainage 12.00 km to 14.00 km, one lane, three Part B items and an invented variant X.

Item	Degree D	Extent E	Weight W (invented)	D × E × W	Maximum, 5 × 5 × W
Crocodile cracks	3	2	8	48	200
Rutting	3	3	4	36	100
Potholes	1	2	8	16	200
Total				100	500

The preliminary index is 100 × (1 − 100 ÷ 500) = 80, which invented variant X reports. Invented variant Y uses the same weights but multiplies the preliminary index by itself and divides by 100, giving 64. Both are correct, and neither can be read without its variant name.

The total cannot exceed the maximum, so the preliminary index runs from 0 (every item at its maximum) to 100 (no distress). Now add a fourth item, edge break, rated absent with an invented weight of 4, so its maximum is 100. The denominator becomes 600 and the preliminary index 83.3, although nothing on the road changed; the item list belongs to the variant.

What does a visual condition index calculation not decide?

Four questions sit outside a visual condition index calculation:

• Which variant and edition apply. The client's terms of reference decide; the version rules printed in TMH 9 Part A give a committee draft no legal standing, so name both in the report. (COTO, 2016b)
• What the number means. The category table belongs to the variant, and Draft TRH 22 cautions that indices should be used with discretion. (CSRA, 1994)
• Whether a segment needs action. Draft TMH 22 (E.1.3) notes that averaging can hide a local problem; Appendix J-8 gives separate example algorithms for classifying needs. (COTO, 2013)
• Other pavement types. Appendix J-6 and ReCAP Part B concern flexible pavements; concrete, block and unpaved roads need their own named methods.

Which misreadings change a visual condition index?

Five misreadings change a visual condition index or its reading:

• "Appendix J-6 continues Addendum A.3." It continues section 6.2.2; TMH 22 did not carry A.3 forward. The insight on visual ratings and the VCI works through A.3 on its own synthetic ratings. (CSRA, 1994) (COTO, 2013)
• "Any weight set with any transform." One variant's weights with another's transform belong to no published variant.
• "An item rated absent can leave the denominator." Dropping it changes the result; above, 83.3 becomes 80.
• "Extent is a percentage." Extent is a class, and TMH 9 Part A (Table A.2) says its percentages are a guideline, not to be interpreted literally. (COTO, 2016b)
• "A change in VCI means a change in the road." A new variant or weight set between surveys creates a step with no change on the road.

What evidence should a visual condition index calculation leave behind?

Keep five records for every segment:

• the original TMH 9 ratings as observed, with road, segment, assessor and date (Part A, Table A.4) and the lane assessed;
• the verification log: checks run, flags, resolutions;
• the named variant with edition and clause, the weight table and any local change;
• each contribution, the denominator, the preliminary and transformed values and the rounding rule;
• the workbook or software version that did the arithmetic.

A one-line method record then lets another analyst reproduce or challenge the number:

VCI [value]; variant [manual, edition, clause]; weight table [identifier, local changes]; ratings [TMH 9 edition, part, form]; rounding [decimal places]; calculated [date, tool version].

The procedure, step by step

1. Confirm which ratings you hold

   Check the TMH 9 edition, part, form and segment; flexible pavements follow Part B. Record: edition, part, form, road, segment, lane, assessor and date.

2. Verify the ratings before any arithmetic

   Run the authority's quality management plan checks and a verification routine such as Draft TMH 22 Appendix J-4. Return a flagged rating to the form or photographs; never edit a number to clear a flag. Record: checks, flags, resolutions.

3. Choose the variant and name it

   Use the variant the client's terms of reference name; if none, choose one published variant, say why and keep it throughout the programme. Quote it fully: VCI (TMH 22 Appendix J-6, Draft March 2013). Record: variant, manual, edition, clause.

4. Fix the item list, the unit and the missing-data rule

   Use the named variant's item list and calculate one index per segment and lane as rated. Zero means assessed and absent, never not looked at. Record: item list, lane and missing-data rule.

5. Take the weights from the named table

   Copy each weight from the named variant's own table, and any default extent for functional items from that variant's text, not from memory or another variant. Record: table identifier and any local change.

6. Weight each rating

   Multiply the degree rating, the extent rating and the weight for each item, using the named variant's own product. Keep the contributions unrounded. Record: every item's contribution.

7. Normalise against the maximum possible

   Divide the contribution total by the variant's maximum total, subtract the ratio from 1 and multiply by 100. An item rated absent stays in the denominator; an item not rated follows the variant's rule. Record: numerator, denominator, preliminary index.

8. Apply the variant's transform, if it has one

   Apply only the transform that belongs to the named variant, exactly as printed in its manual, and round once, at the end. Record: transformed value and rounding rule.

9. Record the method with the number

   Write the variant, weight table, ratings edition and rounding beside every stored or quoted value, because the TMH 18 visual condition files (Draft Version 4, Committee Draft 5.0) have no field for the variant. Record: the one-line method record.

Frequently asked

What is a visual condition index (VCI) for a road pavement?

A visual condition index (VCI) for a road pavement is a single number from 0 to 100 that condenses the degree and extent ratings of many distresses on one assessment segment, so that segments can be compared, ranked and followed over time. The South African road manual TMH 9 supplies the ratings; a named variant, such as Draft TMH 22 Appendix J-6, supplies the formula and weights.

How do you calculate a visual condition index (VCI) from TMH 9 ratings?

Verify the degree and extent ratings in the South African manual TMH 9, then name a published visual condition index (VCI) variant such as Appendix J-6 of Draft TMH 22 (Road Asset Management Manual). Take its weights from its own table, multiply degree, extent and weight for each item, normalise against the maximum total that variant defines, apply that variant's transform if it has one, and record the method beside the number.

Where is the visual condition index (VCI) formula for South African road pavements published?

The South African road manual TMH 9 (Committee Draft Final, May 2016) supplies ratings only. The visual condition index (VCI) calculation is set out in Draft TMH 22 Appendix J-6 (March 2013), in Draft TRH 22 section 6.2.2 (1994), in the proposed modification in its Technical Addendum A.3 and, for low-volume rural roads, in section B.5 of the ReCAP (Research for Community Access Partnership) Part B guideline (Final, September 2020). Each has its own weight table.

Which visual condition index (VCI) variant should a road authority use?

The client's terms of reference decide. TMH 9, the South African manual for visual assessment of road pavements, does not choose between visual condition index (VCI) variants, and the version rules in its Part A give a committee draft no legal standing. Where the terms name none, choose one published variant, record the reason, keep it for the whole programme and quote it by manual, edition and clause every time.

Why must a visual condition index always be quoted with its variant?

The same ratings give different visual condition index (VCI) values for a road pavement under different variants, because item lists, weights, defaults and transforms differ. In the synthetic example, one set of ratings gives 80 under an invented variant X and 64 under an invented variant Y. A VCI without its variant cannot be reproduced, compared with another survey or checked by the next analyst.

Does a visual condition index choose the treatment for a road pavement?

No. A visual condition index (VCI) condenses ratings into one as-is figure and takes no account of traffic or functional classification. Averaging can hide a severe local problem behind better values elsewhere, and the category table that classifies the number belongs to the variant. Needs determination is a separate step, described in Draft TMH 22 (Road Asset Management Manual) Part G and Appendix J-8.

Where this connects

Draft TMH 22 places the VCI among the condition indices in Part E and covers needs determination in Part G; see what TMH 22 covers. TMH 18 carries the value to the receiving authority, and from visual rating to a rehabilitation shortlist follows a rated segment through both manuals. Last reviewed on 21 October 2026 against the editions cited above.

References

1. COTO (2016b). TMH9: Manual for Visual Assessment of Road Pavements. Part A: General. Committee Draft Final, May 2016. Original source.Section A.2: degree and extent
2. COTO (2016j). TMH 9: Manual for Visual Assessment of Road Pavements. Part B: Flexible Pavements. Committee Draft Final, May 2016. Original source.Sections, tables and clauses are cited inline in each guide
3. COTO (2013). Draft TMH 22: Road Asset Management Manual. Draft, March 2013. Original source.Sections, tables and clauses are cited inline in each guide
4. CSRA (1994). Draft TRH22: Pavement Management Systems. Pretoria: Department of Transport, 1994. Original source.Technical Addendum A.3, equations A.3.1–A.3.3 and Table A.3.1, printed pp. 125–126; Section 6.2.2, printed pp. 55–56, for alternative-weight normalisation
5. ReCAP (2020a). Raters' Guideline for Visual Assessment of Low Volume Rural Road Pavements. Part B: Flexible Pavements. Final, September 2020 (GEN 2132A); M. Pinard and R. Geddes, Civil Design Solutions; London: ReCAP (Research for Community Access Partnership) for DFID. Original source.Section B.5 (Visual Condition Index, Tables B.5-1 and B.5-2) and Annex C (calculation of the VCI, including the factor that ignores items not rated)
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

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