Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback

Pavements

Local Calibration of the MEPDG Using Pavement Management Systems

Chapter 2. Three State Selection Approach

Introduction

Based on the 2006 FHWA study eight candidate SHAs were identified as being able to feasibly undertake MEPDG calibration using pavement management system data (FHWA 2006a; FHWA 2006b). However, working with eight SHAs in demonstrating how pavement management system data can be used to calibrate the MEPDG would be very time and cost prohibitive. Therefore, this study narrowed the list of eight SHAs down to three. In order to objectively select three of the eight SHAs, the project team identified the following selection criteria:

  • Availability of data. The selected SHAs should have as complete as possible the data required for MEPDG calibration. This includes both the pavement management system data and the other required data for calibration.
  • Data quality. The availability of high-quality data (correctness, accuracy, reliability, data collection procedures, and quality assurance procedures) is imperative if reliable calibration results are desired.
  • Format of the data. For the convenience of data retrieval (query), relational database like MS Access®, MS SQL Server, or Oracle would be ideal. However, other types of electronic data formats (e.g., MS Excel®) can also be used easily. It will take more effort if the data are in paper archives or in mixed format, if they exist on differing types of computer platforms, or if they use non-compatible referencing systems.
  • Level of data collection effort. There should be enough data to take into account the seasonal variation of pavement responses, enough coverage to contain all typical pavement types, and the data collection approach should be sufficiently standardized so that the developed framework can be easily expended to most other SHAs. Additionally, the data must be stored in a way that allows the performance data to be linked to the specific locations where destructive or nondestructive tests have been taken.
  • The extent of effort required to acquire additional data for the MEPDG calibration. This is related to the data availability and data format mentioned above. If any required data element is missing from the existing data, it will have to be added to the MEPDG calibration database. Depending on the type of missing data, the source of the data available, and the approach needed to re-collect the data, the additional work effort may be significant.
  • Anticipated required IT work for linking various databases. To link the pavement management system data and various types of other data, some IT work will be needed. This may include creating a satellite database, creating the primary keys and foreign keys for relational databases, or combining the data from various sources into one logical database. It is possible that a front-end application may be needed to process and combine the data. The anticipated requirements for each of the SHAs will be considered and rated in terms of significance.
  • Availability of asphalt, concrete, and composite pavements. The selected SHAs should have good coverage of all the three typical pavement types that are within its PMS data collection network. It would be helpful if the system covers different pavement types (e.g., hot mix asphalt [HMA], jointed plain concrete pavement [JPCP]) and construction types (new design or rehabilitation design).
  • Availability of essential data at Level 1 and 2 of MEPDG. Because the availability of Level 1 and Level 2 data increases the accuracy of the resultant calibration, it is desirable to have as much Level 1 and Level 2 data as possible for the key data elements. Most of the eight states involved in the FHWA study showed that the materials data such as the asphalt mix modulus and dynamic modulus of concrete are available at Level 1 (FHWA 2006b). Usually, the traffic volume adjustment factors are Level 3 data.
  • State’s plans to implement the MEPDG. Six of the eight states studied are among the fifteen FHWA Lead States for MEPDG implementation. However, all eight states are active and working to implement the MEPDG within a few years.

The three state selection matrix/criteria proposed and presented herein is designed to permit a realistic assessment of the overall suitability of the eight states to participate in this study. This effort, resulted in the identification of the three SHA that are best suited to contribute to the advancement of the calibration effort nationally using existing pavement management systems.

Selection Concepts

Based on the findings of the two FHWA reports on the use of pavement management system data to calibrate the MEPDG (FHWA 2006a; FHWA 2006b), the research team identified ten primary elements that serve as indicators of a state’s readiness to advance the calibration effort using pavement management systems. These indicators fall neatly into four distinct categories of selection criteria, which are discussed in more detail below. The "readiness indicators" are included in a MS Excel® spreadsheet matrix as elements to be rated by an evaluator. Based on the evaluator’s findings, a rating of 0 to 10 was assigned to each indicator, with 10 representing the most favorable rating or highest degree of conformity with that element required for calibration. Since the selection criteria categories are not necessarily equivalent to each other in terms of qualifying the state’s readiness for calibration, a unique weighting factor has been assigned to each category to reflect their relative importance. For example, the Level of Commitment Category carries a relative weight of 5, whereas the Required Level of Additional Effort Category has been assigned a lesser weight of 3. This permits reflection of the critical importance of a state’s willingness and capacity to dedicate necessary resources to the project as compared to the somewhat less critical indication of the need for additional prep work. Table 1 is a spreadsheet that contains all of the above indicators in a matrix format. It represents one state that has been evaluated for illustration purposes only.

The scoring process calls for the completion of one table for each of the eight SHAs. After the evaluator has assigned a rating of 0 to 10 to each of the indicators, a score is then calculated for each indicator by multiplying the rating by the category’s relative weighting factor. Four separate category scores were determined by subtotaling the individual scores for all indicators in each category. Finally, a single grand total score was computed for each of the eight states by totaling all four category scores. This approach permitted a more focused comparison among states by category as well as by aggregate total score. The sensitivity of this proposed approach can be evaluated by comparing the evaluation conducted by two or more individual evaluators.

Table 1. Selection Criteria Matrix -One State Example.
State XY Evaluator MG Date 3/12/2008
Category Indicators Rating 0 - 10 Weight Factor Guideline/Comments Score (Rating *weight Factor) Cat Score, % Grand total, %
Category I: Level of Commitment State Plan to Implement MEPDG 6 5 If the state has an existing MEPDG implementation plan, a rating of 10 is assigned. If no plan exists, a rating of 0 is assigned. 30 73.3 71
Category I: Level of Commitment Degree of Commitment to Implementation 7 5 If the state is committed to and has a plan to implement the MEPDG & the state is willing and able to dedicate the necessary resources, a rating of 10 is assigned. If the state is unable or unwilling to commit the necessary resources, a rating of 0 is assigned. Otherwise, an intermediate rating is assigned based upon the likelihood of future commitment. 35 73.3 71
Category I: Level of Commitment Evidence of Calibration Activity 9 5 As an indication of the state's commitment to MEPDG implementation, a rating of 10 is assigned if the state has an active calibration effort underway led by a consultant/university or an expert in-house team. If no calibration is underway or planned for the near future, a rating of 0 is assigned. 45 73.3 71
Category II: Availability and Quality of Data Availability of Design and Performance Data (for all pavements) 9 4 If the state can demonstrate the availability of design and performance (distress) data for all 3 pavement types (for new and rehabilitation designs), a rating of 10 is assigned. If data exists for two or only one pavement type, a lesser rating is assigned depending on the availability of data. 36 63.3 71
Category II: Availability and Quality of Data Availability of Essential Data (Materials, Traffic, Construction, Climate, Environment) at Level 1 and/or 2 4 4 If the state can demonstrate the availability of essential calibration data (Materials, Traffic, Construction, Climate, and Environment) at Level 1 and/or Level 2, the state is assigned a rating of 10. If data is only available for some essential data at Level 1 or Level 2 and other data is not available at either of these two levels, the state is assigned a lesser rating depending on the relative amount of data at Levels 1 or 2 in proportion to Level 3 data. 16 63.3 71
Category II: Availability and Quality of Data Data Quality and Objectivity (the state’s opinion regarding their data quality) 6 4 If the state is very confident of their distress data quality and objectivity and demonstrates a solid data QA/QC program, a score of 10 is awarded. Otherwise, the state is assigned a lesser rating depending on their level of confidence in data quality and objectivity. A higher score is awarded to states using automated data collection and analysis technologies. 24 63.3 71
Category III: Required Level of Effort Level of Data Collection Intensity (network vs. project level - frequency of coverage) 7 3 Level of ongoing data collection intensity is evaluated with respect to 1) project/ vs. network level data, 2) frequency of coverage (annually vs. bi- or tri-annually), 3)extent of coverage (data per mile, and 4) level of distress detail (actual measurements - see attached table). Rating is dependent on the degree to which state’s data collection methods conform to the table (10 = all elements met) 21 73.3 71
Category III: Required Level of Effort Anticipated Required IT Work 8 3 If the anticipated IT work required to support local calibration is judged to be none or very little, the state is assigned a rating of 10. If the anticipated IT work required is judged to be moderate, the state is assigned a score of 5; and if the IT work required is judged to be extensive the state is assigned a score of 1 24 73.3 71
Category III: Required Level of Effort Extent of Effort to Acquire Additional Data 7 3 If the extent of effort required to acquire additional data for local calibration is judged to be none or very little, the state is assigned a score of 10. If the extent of effort required is considered to be moderate, the state is assigned a score of 5; and if the extent of effort required is considered to be extensive the state is assigned a score of 1 21 73.3 71
Category IV: Data Format 9 2 If the state pavement management system and other data required for MEPDG calibration are compatible with MS®® Excel , MS Access , or other type of relational format that can be imported (or exported), the state is assigned a rating of 10. Otherwise, the state is assigned a lower rating depending on the availability of acceptable/workable data format. 18 90 71

Discussion of Selection Categories

Following is a more complete description of the selection criteria for each of the four categories.

Category I: Level of Commitment

While this category of measure is not technical in nature, it is arguably one of the most important considerations of all. The successful calibration effort is critically dependent on the willingness and the capacity of the SHA to dedicate the resources (time and financial) necessary to see the project through to fruition. A relative weighting factor of 5 was used to compare the importance of this category against the other categories. It is worth mentioning that the original FHWA reports documented varying levels of commitment between the eight states included in the original exploratory study. The level of commitment therefore needed to be assessed through a rational approach. Thus, the Level of Commitment category was comprised of the following three indicators of a state’s commitment to MEPDG implementation:

  1. State Plan to Implement MEPDG. An existing plan for implementation would be viewed favorably as an indication of a state’s intent to move toward implementation. For this indicator, a rating of 10 is assigned if a SHA has an implementation plan in place. If the state is in the process of working on such a plan, a rating of 5 is assigned. If there are no plans for implementation a rating of 0 is assigned.
  2. Degree of Commitment. This indicator is intended to provide some measure of a SHAs willingness to fully participate in the effort and its capacity to dedicate the resources needed for MEPDG calibration. While relatively subjective in nature, this rating would shed light on the subject of commitment from the perspective of the responding SHA representatives.
  3. Evidence of Calibration Activity. On-going efforts by a SHA to calibrate models were interpreted as a positive indicator of a state’s commitment to MEPDG implementation. A higher rating was assigned where evidence of calibration activity by a consultant, university or expert in-house team was demonstrated. It was felt by the contract team that a contractual commitment to the calibration activity would imply a strong desire to get it done within a specified time frame as part of a larger implementation plan.

Category II: Availability of Data

Clearly, the importance of data, complete to the extent possible, cannot be overstated. Design, materials, construction, performance histories, traffic and environmental data at Level 1 and/or Level 2 are essential for successful model calibration. Therefore this category, which carries a relative weighting factor of 4, is comprised of the following two data indicators:

  1. Availability of Design and Performance Data. The overall MEPDG implementation effort will eventually require models to be calibrated for all pavement surface types (flexible, rigid, and composite) for both new and rehabilitation projects. Therefore, the availability of both design and performance (distress) data for different projects is considered to be a key indicator of a SHAs preparedness for calibration and eventual implementation.
  2. Availability of Essential Data at Level 1 and or Level 2. Materials, traffic, construction and environment data is essential for MEPDG models calibration. The more of this type of available data that conforms to Levels 1 and 2, the less variability is expected in the design output. Therefore, a SHA was assigned a higher rating for this indicator if it was able to demonstrate the availability of a high percentage (relative to the other seven states) of this data that conforms to Levels 1 and/or 2.
  3. Data Quality and Objectivity. Based on information provided by the appropriate SHA representative, this provided an indication of the level of confidence the state has in the quality of its pavement management data. Given the enormity of work involved with objectively determining data quality, information supporting these criteria was based to a large extent on the state’s own experience and the opinions of its representatives.

Category III: Required Level of Effort

Category III is included in the selection matrix to capture some understanding of the magnitude of additional work (such as supplementing the existing pavement management system data and performing any related IT work) the candidate states would have had to undertake in support of the calibration effort. The Required Level of Effort Category was designed as an attempt to measure the readiness of the states to move forward with calibration in terms of the compatibility of their existing pavement management data, additional data needs to be collected, and IT architecture required for calibration. A relative weighting factor of 3 has been assigned to this category, which is comprised specifically of the following indicators:

  1. Level of Data Collection Intensity. This is intended to provide an indication of the suitability of the state’s on-going distress data collection activities with regard to project vs. network-level coverage, frequency of condition surveys, extent of coverage in terms of survey sample size, and the degree to which the collected distress data conforms with MEPDG model calibration requirements presented in table 2.
  2. Anticipated Required IT Work. This indicator serves to gauge the magnitude of additional IT work above and beyond existing capabilities that would be required to minimally accommodate calibration activities including data linkage and creation of keys for relational databases.
  3. Extent of Effort to Acquire Additional Data. This indicator is needed to gauge the amount of additional work required to add any missing data elements. Consideration should be given to the type of missing data and the extent of work that would be required to capture or re-capture that data.

Table 2. MEPDG Required Distresses for Local Calibration.

MEPDG Required Distresses for Local Calibration
HMA Distress Data
IRI1 in/mile
Asphalt top/down (longitudinal) cracking ft/mile
Asphalt bottom/up (alligator) cracking % cracked
per section length
Low temperature thermal cracking
(transverse)
ft/mile
Asphalt rutting2 (permanent deformation) inches
JPCP Distress Data
IRI1 in/mile
Transverse cracking ft/mile
% slab cracked per section
Mean joint faulting2 inches
Continuously Reinforced Concrete Pavement (CRCP) Distress Data
IRI1 in/mile
Number of punchouts per/mile
Maximum crack width in
Minimum crack load transfer (transverse) LTE%
Minimum crack spacing ft
Maximum crack spacing ft

1International Roughness Index, typically measured every tenth of a mile

2Average, standard deviation, COV, maximum, minimum

Category IV: Data Format

The final category attempts to provide an understanding of the degree of ease with which the necessary data may be manipulated (i.e., relational format). A relative weighting factor of 2 is used to compare the importance of this category against the other three categories. The following indicator comprises Category IV:

  1. Data Format. State pavement management data in a relational format such as MS Excel® or MS Access® was viewed as a positive indicator of easy manipulation.

Other Selection Considerations

From the perspective of this study and to generate the maximum benefit possible to the greatest number of states, it was desirable to include those states that are the most representative of the "typical" highway agency. Stated differently, the ultimate selection of the most advanced or mature agency with regard to the status of their MEPDG implementation efforts would not necessarily yield great benefits to a less mature state with an earnest desire to move forward with implementation. For this reason, the selection matrix presented above was augmented with extensive discussion among the research team regarding the advantages and disadvantages of including each state in the study. For example, discussion topics included concerns raised by the team regarding the capacity of a state’s pavement management system to objectively support the calibration effort. Or, perhaps the advantages of a particular agency’s state-of-the-art distress data collection and analysis methods outweighed some other identified weakness in their pavement management system.

Once the research team reached consensus with regard to the scores of all eight agencies, the three states with scores that most closely approximate the median score for the entire group were selected for inclusion in the study. Based upon the statistical "spread" of the resulting scores, the research team selected a group of three states with scores slightly above or slightly below the median score (i.e., states # 3, 4, and 5 instead of # 4, 5, and 6).

Summary

The described selection criteria provided a rational approach to evaluate the suitability of the eight states in moving forward with MEPDG calibration using data contained within a pavement management system. The selected criteria not only provides an assessment of data availability, data storage format, and data accessibility, but also the willingness and availability of SHA staff to conduct the level of effort needed in the MEPDG calibration process. In addition, selection of the three potential SHAs also included consideration to maximizing the study outcomes by selecting states that represent the "typical" highway agency.

More Information

Contact

Nastaran Saadatmand
Office of Asset Management, Pavements, and Construction
202-366-1337
E-mail Nastaran

 
 
Updated: 10/12/2011
 

FHWA
United States Department of Transportation - Federal Highway Administration