Direct Enforcement in Belgium with High Speed Weigh-in-Motion (HS-WIM)

Round 1

Reviewer 1 Report

Interesting and timely topic. The experience of Wallonia in WIM and direct enforcement is very valuable for all the end users, in Europe and worldwide.

However, the paper needs quite significant improvements before to be published.

Abstract: replace the word "security" by "safety" (line 3).

Keywords: add "Enforcement" or "Direct Enforcement".

1. Introduction

line 18: replace "tested" by "checked".

line 20: replace "number of weighing operations" by "number of static weighings".

line 23: replace "to become a legal metrology" by "to be certified by the legal metrology".

line 24: Explain the relation between the content and objectives of the paper and the multiple sensor WIM (MS-WIM) described in the reference [3]. MS-WIM does not cover WIM with 2 lines of sensors as used in Wallonia. Moreover the sentence: "The purpose of this theory is to reduce the total amount of dynamical load for vehicles moving at speeds" is irrelevant. MS-WIM does not affect the dynamics of hte vehicles, but allows a WIM system to average the variations of the dynamic axle loads along the road, and to more accurately estimate the static loads.

lines 33-34: the sentence "The article simulates the dynamic load as a side effect of weight measurement for direct enforcement" is not true. No simulation results are reported in the paper, but only real measurements.

line 37: replace the word "campaign", because the renewal of the roadway improves the quality of ALL the measurements, not only of one campaign.

lines 37-38: pease refer to the right section of the paper, where the conditions are detailed, after the sentence "Under some individual validity conditions for vehicles". Then, classes A and C are valid only for axle loads in the OIML R-134, not for gross vehicle weights (which are classified in classes 2, 5, 10, etc.). Please amend accordingly.

line 40: replace "guaranteed accuracy" by a more appropriate term. The accuracy should meet some performances but there is no real guarantee.

lines 41-42: "an entire new road" should be reworded. Only a road section of a few meters (hundreds of meters) was renewed. It may be said that all the pavement layers were replaced along a given distance.

line 42: complement "a double grid of 2 sensors each". Replace "good parameters" by something more precise. What is good?

2. Materials and Methods

2.1. Road and sensors layout description

line 48: replace "this got the HS-WIM in the class II" by "this make the WIM site in the class II".

line 54: replace "implemented" by "installed".

line 55: replace "figure 3" by "figure 1".

line 56: replace "piezo-quarz beams" by "piezo-quarz sensors" or by "piezo-quarz bars".

lines 56-57: Should you replace "pressure sensors" by "magnetic loops"?

line 58: Should you replace "Two independent sets of sensors" by "Two independent WIM systems"? There are also two independent electronics, with their own algoritms...

line 59 (and later in the paper): the word "prediction of the weight" seems not adequate. It is suggested to use "weight estimate" or "weight measurement", likely complemented by "to be used for enforcement". The second measurement should be named "check measurement".

2.2. Data collection

line 66: repalce "following type of vehicles" by "following types of vehicles".

line 67: replace "semi-trailer with two-axle rigid vehicle and a three-axle daw-bar trailer" by "a two-axle tractor with a semi-trailer on a tridem axle".

line 68: delete "with" and replace "two-axle rigid vehicle" by "i.e. two-axle rigid vehicles with a maximum authorized weight of 3.5 tons".

lines 69-70: replace "dynamic weights" by "weigts measured in motion (at speed)". Replace the word "prediction" (see above). Replace "further away" by "downstream".

lines 71-74: there is some confusion between the value 1 to validate measurements (in the text) and the wording OFF/ON two lines below and in Table 1. Please harmonize. The wording "regular passing" seems strange. Please amend. The criteria of validation should be more precisely described. Are they sufficient?

lines 76-77: "Other times, they deviated the whole traffic of T2S3...". Do the control staff operate 24/24 and 7/7? If not the sentence should be reworded.

Additional comments/questions on this section: it would be good to give the proportion of the whole traffic reported in Table 1, i.e. which proportion of the total number of T2S3 which passed on the motorway all along the 6-month period, is represented by the 404/180 vehicles in the table 1? And how were the measured vehicles chosen among the whole traffic? How many days/hours of measurements were carried out along the whole period? Did the measurements target the fully loaded/overloaded vehicles or did they contain also empty and half-loaded vehicles? A histogram of the reference measured gross weights would be welcome.

3. Material and method

The title of this section is identical to the title of the previous section 2. Either, transform this section into a sub-section of section 2, as 2.3 "Relative errors of the measurements", or rename this section. Harmonize capitals and no capital letters in the section and sub-section titles.

line 78: replace "machine" by "instrument", and "dynamic weight" by "weight measured in motion" (not "computed with the HS-WIM system", but "measured by..."). Replace the word "prediction" (see above).

line 82: replace "a lower bound" by "a sample estimate". Delete "is calculated,". Replace the word "prediction" (see above).

4. Results

lines 85-87: the tables 1 to 2 give results for gross vehicle weights and second axle loads of the T2S3 in a first part, and other results below. For the U2, there are two lines in the table 3, one for GVW and one for single axles. The first sentence must be revised to better explain why these entities are differentiated, and how the results were used.

lines 88-92: replace "experimental" by "sample". Please define what is pi0 and delta0 (it seems according to the COST323 that delta 0 dos not exist, likely delta min? Please explain). The last sentence is not understandable and must be reworded.

lines 97-98: as above, classes A and C only deals with axle load.

line 101: replace "COST report" by "COST specifications".

Tables 1, 2 and 3 should be placed one after the other (do not isolate table 3).

The concept and above all the wording of confusion matrix is not  clear. The authors should introduce the two risks: (1) missing an overload by the HS-WIM system, and (2) wrongly identifying overload, which may lead to a wrong fine in case of direct enforcement. Then, it should be emphasized that the risk 2 is much worst as the risk 1.  In addition, for each sample of measurements, both risks should be presented. As it is, the confusion matrix is not clear as well as the discussion. 

It may be underlined that the risk 2 is null in the presented results, while the risk 1 is quite high. Should the user reduce the threshold of detection/penalization, or recalibrate the WIM system?

lines 107 and 111: table numbers are missing; "51" should be replaced by "56".

5. Discussion

This section is rather poor and below the expectations. The only finding is that filtering the results increases the accuracy; quite obvious. The authors have more findings to develop, above all the efficiency of the certified WIM system for direct enforcement. They should rewrite the discussion, showing that the systems meets the requirements for direct enforcement and open the way to a new and much more efficient control process. We can also expect some information about the algorithms and the criteria of validation implemented in the WIM system, which both highly contribute to its performances. That would be the most original and useful information for the scientific community and the end users.

6. Conclusion

Again, as explained in the previous §, the conclusion cannot only focus on the benefit of rebuilding the road. The contribution of the algorithms of the WIM systems should be emphasized. Another question needs to be discussed and analyzed in the previous sections: because the risk 2 (wrong overload detection) is null while the risk 1 (missing overload detection) is rather high, the threshold of overloading or the calibration should be discussed, and a sensitivity study carried out.

In summary, the work reported in this paper is really interesting and useful, but the paper needs some significant improvement and complements (which surely exist in the collected data) to be published in an International journal.  The English language needs significant improvements.

However, after a significant revision and some complements and developments, which may be derived from the available data, the paper could be very good and valuable.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

From my point of view, the paper cannot be published in this form.

There are more major observations:

- Table 1: ON/OFF for T2S3 vehicles means “The validity criterion is set to ON if the passing is valid, and to OFF otherwise”. What means ‘/’ for U2 vehicles?

- Typo: “In the tables ?? and ??”, Replace question mark with numbers.

- Major (very major) observation: Detecting the weight of vehicles through the HS-WIM system is not a new problem.  There are studies and systems that solve this. Even on the website of the International Society for Weight in Motion (ISWIM) there are mentioned solutions for HS WIM (http://www.is-wim.org/index.php%3Fnm%3D2%26nsm%3D1%26lg%3Den) . So, in the Introduction section, you must present the objectives of the paper, what it proposes to develop in relation to the existing HS-WIM solutions. For example, B. Jacob (the author you quote in the paper with a study from 2008) highlighted in 2010 (in the article "Improving truck safety: Potential of weigh-in-motion technology") the advantages and technologies of HS-WIM - so it is no longer a novelty the fact that it is more efficient than LS-HIM. Present in the introduction what you add new through your work.

- Major observation: What I understood is that the paper proposal validates the HS-WIM solution proposed in reference [7]  through an analysis of the results obtained by this system compared to an LS-WIM system. Correct? In this case highlight this contribution and this objective in the Introduction. However, I would like to specify that it does not seem to me sufficient for a paper published in a journal to only validate a solution presented in another article that was probably highlighted there. It would have been better to present any improvements you can make to that solution - for example to increase accuracy. Did you bring something like this to this paper proposal? If so, highlight this - it would be an important contribution.

- Major observation: In introduction you said: “The whole system consists of an entirely new road and a double grid of sensors with good parameters for the weighing algorithms” and in Results:  “We observe that the algorithm used by this HS-WIM system display “ . Do you elaborate an algorithm? If yes, present it!

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors responded to my observations. The contribution made by the authors is now clear. I have no other observations.