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


Local Calibration of the MEPDG Using Pavement Management Systems

Chapter 4. Single State Selection


In preparing for discussions with the three States (Mississippi, North Carolina, and Minnesota), the project staff sought to confirm the information used during the initial evaluation of the eight States. It was readily recognized that the study done by Hudson et al. was conducted in 2006, and in the two years since that study many things may have changed. For example, several States have been actively involved in evaluation and implementation studies for the MEPDG. The research team sought to explore the current status of work underway in the State as pertaining to the MEPDG implementation, the availability of information in the pavement management database to support the calibration efforts, and the potential level of support that may exist for testing a proposed framework for the use of available State data in calibrating the MEPDG. Specifically, the following factors were considered:

  • Level of support in terms of staff requirements.
  • Staff availability.
  • The State’s level of support in terms of budget.
  • The computer hardware and operating systems used for related pavement databases.
  • The level of IT and database skills the State.
  • Level of commitment to this effort by upper management.
  • Likelihood of success with the implementation.

The meetings in each of the three States were informal, and were completed in about one and a half business days. The format generally included meeting with all interested personnel to introduce the research effort and the people involved. State agency representatives were asked to discuss, in general terms, the status of their MEPDG implementation efforts. Any university studies that may have been conducted, or are currently under way were discussed, to gain an understanding of the objectives, results, and current status of the work. Discussions were held with representatives of Design, Traffic, and Materials to understand the status of implementation preparedness in each of these areas. Discussions were held with representatives of the Pavement Management group, to assess the availability and format of required data, and the level of effort generally required to access this information.

The likelihood of success with the implementation was primarily gauged by the enthusiasm the State exhibited for the effort, and the existence of a plan for continuation of the effort. Project staff listened for indications of support by upper management, and where necessary queried meeting participants as to the degree of support they received for their efforts.

State Visit Summaries

Following are observations and comments from the meetings with the three State highway agencies.

Minnesota Department of Transportation

Meetings were held with the Minnesota Department of Transportation (Mn/DOT) on Sept, 2008. State participants in the meeting included Mr. David Janisch (Pavement Management), Jerry Geib (Pavement Design), Matt Oman (Traffic), and Curt Turgeon (Pavement Engineer). On Day 2 we were joined by Ms. Maureen Jensen, who had worked extensively in the evaluation and implementation of the MEPDG prior to moving to another area in the Department.

Discussions generally confirmed the advantages and disadvantages observed from the earlier work. While Mn/DOT was highly committed to the effort previously, they exhausted their implementation budget while finding a number of apparent problems in the software, and at this point are waiting to take further action once the software is in a more stable position. While the Department continues to collect a great deal of distress and roughness data, the format of the data is not consistent with MEPDG predictions, meaning that the effort required to do meaningful comparisons would be fairly high. While there have been increases in the amount of traffic data collected, staffing shortages have prevented the management and manipulation of data needed to produce the required load spectra information or truck weight road groups.

In general, Mn/DOT feels it has a design process in place that provides an acceptable result, with expected life and actual life in close agreement. As a result there is a lack of justification for making a major change. At this point the cost would be great, with little perceived benefit.

Much of the Mn/DOT implementation work was done in concert with Dr. Lev Kazanovich and the University of Minnesota. An attempt was made to meet with Dr. Kazanovich, but unfortunately he was out of the country during the time of our visit.

Mississippi Department of Transportation

Meetings were held with the Mississippi Department of Transportation (MSDOT) on October 20-21, 2008. State participants in the meetings included Mr. Bill Barstis (Research), Ms. Cindy Drake (Pavement Management), Mr. Jeff Wages (Construction and Materials), and Mr. Trung Trinh (Traffic). Mr. Roger McWilliams of the Division FHWA office also joined us during the meetings. Discussions with Mississippi DOT personnel confirmed the previous findings, and indicated a great deal of advancement in some areas of implementation readiness in the intervening time period. MSDOT sponsored a series of twelve "support studies" as part of their implementation efforts. About half of these involved materials characterizations, which are nearly complete for typical materials used in the State. In these areas they are building libraries of typical material properties for design use.

The State pavement management system is well populated with time series performance monitoring data, and much of that is in the proper format for comparison with MEPDG predictions. Unfortunately the disadvantages previously noted, including a direct link between sections in the pavement management system and material properties still exist, and Mr. Wages reported a great deal of time and effort have gone into locating construction records and obtaining material properties needed from those records. Therefore, it is still believed that the levels of effort necessary to gather required information will be relatively high. It appears that this may be a fairly common characteristic among most State highway agencies.

One thing that impressed the research team during the meetings with MSDOT was the motivation obvious in the implementation support, and the eagerness to meet and work with the team. Mr. Bill Barstis and Ms. Cindy Drake both expressed a desire to participate in the development of the framework, as they felt it would benefit their, and other, implementation efforts. Mr. Barstis indicated that management had generally been very supportive of the efforts thus far, and he felt that they would continue to be supportive as long as forward progress was being made.

In discussing traffic inputs, the research team found that Mississippi DOT is collecting a great deal of traffic data, and had adequate data management facilities in place. However, they had not begun the process of establishing Truck Weight Road Groups or Truck Traffic Classifications yet. References were provided for States that had done this work, in hopes that they could help Mr. Trinh in developing an approach to complete this effort.

North Carolina Department of Transportation

Meetings were held with the NCDOT on October 28-29, 2008. State participants at the initial meeting were Ms. Judith Corley-Lay and Mr. Neil Mastin. The research team had an opportunity to visit with Mr. Clark Morrison (Pavement Design), Mr. Kent Taylor (Traffic), and Mr. Jack Cowsert (Materials) at a later time. Mr. Jim Phelps of the North Carolina FHWA Division office joined us for all of the discussions. The research team learned that NCDOT has a number of support projects under way, generally through Dr. Richard Kim at North Carolina State University. Projects are under way to develop Dynamic Modulus values for typical NCDOT asphalt mixtures, to investigate traffic data status and needs, and to look at statewide calibration needs. All of these projects indicate a fairly high degree of continued interest in the implementation of the MEPDG. Of specific interest was the statewide calibration study, begun in 2007 and scheduled to be completed in August 2009. This study specifically sought to use pavement management and other data for calibration, and found that many estimations and correlations had to be used due to lack of sufficient information.

One important lesson learned during the meetings was that most of the data stored in the NCDOT pavement management database is referenced by County Route and milepoint, meaning that location referencing may be extremely difficult. This may make information location and retrieval nearly impossible, even before the format inconsistencies are considered. Distress data are not stored in a manner consistent with MEPDG predictions, meaning that it will be very difficult to directly relate the two. Still, NCDOT has completed some studies looking at existing roads and MEPDG predictions, and found that performance predictions are poor. They hope that national studies and model improvements as a part of NCHRP 1-40 studies will improve predictive capabilities.

Discussions revealed that while there have been advancements in the area of material characterization, and a study is under way with Dr. Kim on State calibration, it will be difficult to use available information given the current location referencing method and data format. Still, NCDOT remains committed to the effort, and seems genuinely eager to assist if possible.

State Recommended for the Study

The project staff very much appreciated the willingness of DOT staff in all three States to sit and discuss their efforts with the research team. All three of the States exhibited a great deal of interest in the success of this effort. It was clearly obvious that all three States had been deeply involved in the implementation activities for some time, and were well aware of the input needs, and difficulties in developing some of the requirements.

The research team learned something different from each of the State discussions because each of them offered a different perspective on the same problem. Based on our assessment, the NCDOT was recommended and accepted for participation in this study. The NCDOT recommendation was based on the following:

  • NCDOT personnel have expressed interest and enthusiasm for the project. They have previously initiated activities in this area that will benefit the project team.
  • The Pavement Management Unit is willing to commit engineers to work with our team to populate the MEPDG calibration database required for calibration.
  • NCDOT has performed much of the material testing required for Level 1 and 2 data inputs (in particular dynamic modulus values for typical asphalt mixes) and has the resulting data available electronically. The Traffic Surveys group is actively pursuing a research project for determining higher level MEPDG traffic needs.
  • In general, all Level 2 inputs can be populated with existing data. Since material sources and suppliers will not be known at the design stage of a project, NCDOT has stated an interest in moving forward with calibration to Level 2 inputs only.
  • The Department has the AgileAssets pavement management software in place to provide inventory and performance data needed to calibrate the MEPDG performance prediction models. Pavement deterioration models are in place.
  • Data on two pavement types (HMA and PCC) are available at a variety of traffic volumes.

There are several challenges that need to be addressed to successfully test the proposed framework. The challenges identified by the team, and proposed strategies for addressing these challenges, are provided below:

  • If IT involvement is required, NCDOT stated that they would not be able to meet the proposed timeframe for completing the project. Although not required under the contract, one of the APTech team members (Stantec) developed a preliminary version of the MEPDG calibration database that stores the inputs outlined in the Preliminary Framework. Because of the availability of this database, NCDOT would be able to populate it using existing data sources without requiring intervention from IT.
  • During the interview with NCDOT personnel, some concerns arose because the pavement management database references data by County Route and milepost, which was expected to cause some problems with location referencing. This issue was addressed by selecting specific sections that were used during the calibration study. NCDOT has already identified 10 HMA projects that were initially constructed in 1995 for which they have construction documents and maintenance histories available. Through the use of known locations, the referencing issues are minimized. Although each of these projects is an asphalt pavement, Ms. Judith Corley-Lay, NCDOT Pavement Engineer, has agreed to identify similar sites for the other surface types that were calibrated during the study.
  • NCDOT pavement condition surveys do not conform to the LTPP Distress Identification Manual(Miller and Bellinger 2003), which is typical of the majority of state DOTs. As a result, the research team conducted calibration efforts based on the performance data contained within the NCDOT pavement management system and provide discussion on the impacts, if any, of using non-LTPP defined pavement condition distress on the MEPDG prediction models. The findings from this assessment will be useful to other states faced with similar historic pavement condition surveys that are not based on the LTPP Distress Identification Manual.

More Information


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

Updated: 10/12/2011

United States Department of Transportation - Federal Highway Administration