This paper was prepared for the 4th International Conference on Managing Pavements, Durban, South Africa, May 1998

DEVELOPING IOWA'S STATEWIDE PAVEMENT MANAGEMENT PROGRAM

Omar G. Smadi and Tom H. Maze
Center for Transportation Research and Education
Iowa State University

Abstract

The state of Iowa has embarked on an effort to develop a statewide pavement management system (PMS). The project, the Iowa Pavement Management Program (IPMP), will cover 38,000 km (23,500 miles) of roads operated under three levels of government (state, county, and city). The mission of the project is to support both project-level and network-level PMS conducted by local (city and county) and regional governments and the Iowa Department of Transportation. This paper discusses the technical aspects of pavement condition data collection, development of a dynamically segmented database, and calibration of pavement management software for both project- and network-level analyses. Finally, training will be provided for all transportation agencies interested in using the data and software for the management of pavements within their own jurisdictions.

1. INTRODUCTION

In 1994 the state of Iowa embarked on an effort to develop a database and provide a set of tools to allow the state and regional, county, and city governments to perform pavement management on the roadways under each agency's jurisdiction. This project was titled the Iowa Pavement Management Program (IPMP) and covers roughly 38,000 km (23,500 miles) of highways under the operation of state, county, and city governments.

The IPMP divided the activities required to support pavement management into four fundamental elements. They involve 1) constructing a database of data elements needed to conduct pavement management, 2) collecting the pavement history and condition data to populate the database, 3) making available the decision support tools to conduct pavement management at each level of government (local, regional, or statewide), and 4) making pavement condition and history data and pavement management system parameters (e.g., performance curves, pavement treatment costs, values of minimum pavement condition to trigger the application of a treatment, etc.) available to individual jurisdictions. With these four elements, Iowa governmental agencies may perform pavement management within their own jurisdictions.

We believe that these four supporting elements may be better and more economically conducted on a statewide basis than on a local basis. Significant economies to scale are available by jointly conducting processes that are common to all jurisdictions. For example, on a statewide basis only one contract is negotiated with a pavement condition data collection contractor to collect condition data for the entire state. On the other hand, if each agency had to contract for services on its own, the sheer magnitude of the number of agreements to be executed would make statewide data collection problematic and costly (there are 99 counties in Iowa alone, and an even larger number of cities would be involved). However, the actual management of pavements is left to the managers and engineers at each level of government. These individuals have the specific knowledge of the local pavement network and local resource programming issues and are better equipped to apply local knowledge to resource programming decisions. Further, the authority to direct resources at the local level is uniformly and jealously guarded by local governmental agencies throughout the state. Ultimately, this division of pavement management activities between a statewide consortium of highway operating agencies (cities, counties, and the state) and regional governments allows highway operating agencies to:

Have ready access to high quality pavement data that are consistent across all jurisdictions and retain programming authority over pavement resources within their own jurisdictions.

Because of the scale and the large number of jurisdictions involved in the development of the IPMP, the planning, design, and development process has taken nearly three years to unfold. The three-year process was sequential and, because of the large number of jurisdictions involved, required a great deal of consensus building to generate trust among the various levels of government involved. Further, because of the large number of jurisdictions, a great deal of time was required to plan and design the database just to develop a pavement inventory and history structure that would accommodate each jurisdiction's needs.

2. BACKGROUND

Initially the IPMP was formed to meet the requirements of the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991. ISTEA mandated six (6) transportation management systems and a traffic monitoring system (in reality, a seventh management system) in each state. One of the management systems mandated by ISTEA was a pavement management system. ISTEA required that all federal-aid-eligible (FAE) highways in the state be "covered" by a pavement management system. Later, the Federal Highway Administration developed rules defining what the pavement management system was to include. In general, the federal guidelines left a generous amount of flexibility in the specific design of the system and the path chosen to satisfy the statewide requirement.

In Iowa, part of the FAE highway network is under the management of the Iowa Department of Transportation (Iowa DOT), that is, the state, while the remainder is under the management of county and city governments. Further, because regional and metropolitan councils of government program federal funds at the metropolitan or regional level, a fourth layer of government is involved. Thus, developing the IPMP and gaining support for participation in the IPMP required developing support from all four constituent groups.

In Iowa, the FAE highway network covers about 43,000 km (27,000 miles), with 16,000 km (10,000 miles) under the jurisdiction of the state, 21,500 km (13,500 miles) under the jurisdiction of counties, and 5,500 km (3,500 miles) under the jurisdiction of cities. Out of the state's portion of the FAE highway network, about 5,600 km (3,500 miles) are on the National Highway System (NHS). Federal funding for NHS is programmed separately from non-NHS highways and, therefore, the NHS was not included in the IPMP but covered in a separate pavement management system.

All states were left to develop their own unique approach to implementing pavement management, and Iowa's approach tends to be one of the more centralized approaches. More specifically, data collection, database development, and pavement management software calibration are activities each agency would have to do if it were going to develop a pavement management system for its own jurisdiction. These are common activities, and by conducting them on a statewide basis it was believed that significant resources could be saved through economies to scale. Centrally, the state of Iowa could afford to make greater pavement management expertise available than individual agencies could afford on their own. Centrally, the state of Iowa could procure data collection and database services to provide accurate data and data that are uniform across all jurisdictions in the state. Further, the state could guarantee that all four levels of government would have sufficient data and tools to perform their own pavement management.

The National Highway System Designation Act of 1995 made the implementation of the pavement management system optional. Iowa's Transportation Management Systems Policy Committee, which oversees the development and implementation of all of Iowa's management systems, decided to continue the development of the statewide pavement management effort even without the mandate, but allowed agencies the option of not participating. The decision whether to participate was to be made on a regional basis. In Iowa there are eight metropolitan planning offices (MPOs) and 18 regional planning affiliations (RPAs) located totally or partially within Iowa. (An RPA is a planning organization of several cities and counties that serves a rural area just as an MPO serves an urban area.) The decision to participate in the statewide management system is made at the RPA and MPO level. Thus, if an RPA or MPO votes to participate in the IPMP, then all the cities and counties in the region are also participating in IPMP, and if the RPA or MPO votes not to participate, then the cities and counties in the region are excluded from participation. Although some of the RPAs and MPOs are still considering whether to participate in the IPMP, the majority have decided to continue even though their participation is not mandated.

The IPMP started in 1994 and is still in the implementation phase. Full system operation will start in early 1998. After full operation of the IPMP, training will be provided for local and regional governmental agencies to help them utilize all of the system capabilities.

3. PROJECT DESCRIPTION

Nothing in the development of a statewide system for collecting and disseminating pavement management data and the development of tools for use in pavement management presents an insurmountable technical challenge. However, there are several nontechnical institutional issues that did and still do present significant challenges. We reported on these nontechnical challenges elsewhere (Maze and Smadi, 1996). It is important to note, however, that a significant level of effort is required to overcome nontechnical issues. For example, significant resources were devoted to developing a mission statement, goals, and objectives and to presenting implementation plans at state-level meetings of regional planners, county engineers, and city public works officials and through print media. These efforts were made so that all individuals with an interest in the allocation of highway resources would understand and trust the process being employed. Further, to make sure that local and regional governments did not perceive that the system would be used by the Iowa DOT to influence the distribution of state highway funding, the centralized system was to be developed and maintained by a disinterested third party, the Center for Transportation Research and Education (CTRE) at Iowa State University.

Representatives from all the involved governments and the Federal Highway Administration formed a non-NHS task force to direct the planning, development, and implementation of the IPMP, while CTRE provides the staff to support the task force. Early in the development of the IPMP, while the mission, goals, and objectives were being developed and while plans were being developed, the task force met about once a month. Once the plans began to be implemented, the task force met less frequently and is now (spring 1997) meeting only on an exception basis. So far, the task force has established three committees to supervise three major tasks. The data collection committee is responsible for the evaluation and selection of an automated distress vendor to collect pavement distress data. The database committee is responsible for the development of the IPMP geographic information system (GIS) database. Finally, the pavement management software committee is responsible for the evaluation, selection, and calibration of the pavement management tools selected for the IPMP.

4. PROJECT PHASES

The IPMP project is divided into three major phases. The first phase of the project (Design Phase) has been completed. Work is being conducted now on the second phase (Implementation Phase), and it will be completed by the end of 1997. The third phase (Operation Phase) will start in early 1998 and will be a continuing activity of training, system evaluation, database maintenance, and calibration of the pavement management tools.

Each phase consists of several tasks. Each task has a time line to be completed, specific activities to be conducted, and recommendations for future considerations. Following is a brief description of the tasks under each phase of the IPMP project.

4.1 Design Phase (Phase I)

The design phase main objective was to provide guidelines for the design of the statewide pavement management system. During the design phase, issues considered included data needs, database and pavement management tools, information exchange, and system government and support. A very purposeful attempt was made during the design stage to distribute as much information about the project to regional and local governments as possible. After completing each task under the design phase, a set of recommendations was established and future directions for implementation were determined.

Decisions made during the design phase were related to the performance and technical specifications of the IPMP. The more important decisions included the decision to collect condition data, conduct database development, and populate the database centrally rather than distribute the process among the local and regional governments. It was also decided that through the IPMP the referencing system would migrate from legacy linear referencing systems to real world coordinates (i.e., latitude and longitude).

4.2 Implementation Phase (Phase II)

The implementation phase objective was to develop the various components of the IPMP, test and evaluate the components, and finally develop a plan for operating the system. The implementation tasks included the following:

  • establish the pavement management computer database
  • implement the systems for the physical inventory of pavement
  • evaluate financial, institutional, and technological pavement condition data collections options
  • collect pavement construction and cost historical data from state and local agencies
  • evaluate and select pavement management analysis software
  • collect pavement condition base line data
  • calibrate pavement management analysis system and implement database
  • develop data exchange mechanism between the central database and state, regional, and local users
  • test and evaluate system
  • evaluate system at high level, comparing system to original system objectives

At the time of this writing, implementation of the system is being completed. The decision was made to collect the baseline pavement condition data over a two-year period, half the network during each year. In 1996 the first half was collected, and the second half will be collected during 1997. However, a GIS database has been established for all pavement management related data, the physical inventory system has been built, a decision was reached on an automated pavement condition data collection system, and baseline condition data are being collected for half of the FAE highway system. Much work has been conducted on selecting pavement management analysis software, and specific packages will be selected shortly. The remaining tasks will be completed during the remainder of 1997.

We see the task of developing a data exchange mechanism for pavement condition data to be most challenging. Some local and regional governments are likely to want the data in paper copies or in simple text files with literal descriptions for data collection points (as opposed to latitude and longitude). Others will want data for their jurisdiction processed and ready to be uploaded in the pavement management analysis software. Others may want the data in a format compatible with their in-house GIS, while still others may simply want to use the software platform we are using to house the statewide database. Another option for data exchange we intend to make available is the ability to obtain the pavement management data for a city, county, or region in a format compatible with an inexpensive desktop GIS software with a standard set of data queries already preprogrammed, thus becoming an executive management information system. Other transportation infrastructure will be integrated in the executive information, including data from other transportation infrastructure management systems (e.g., the bridge management system) and from other infrastructure funding and programming systems (e.g., the statewide transportation improvement program, the highway sufficiency rating system, and the five-year transportation improvement program) to allow systemwide resource allocation decision making.

4.3 Operation Phase (Phase III)

The final phase is the ongoing operation of the IPMP and involves the training and support of staff members from local and regional governmental agencies and the Iowa DOT on the use of the IPMP tools and on pavement management in general. Also, system testing, evaluation, and updating will be conducted on regular intervals to ensure proper operation of the system. The operation phase also involves maintenance of the system by continually adding to the pavement condition database, maintaining the database, and distributing historical and the most recently collected pavement condition data. In the future, the IPMP task force will decide on an organization to maintain and house the ongoing operation and management of the statewide pavement management system.

5. MAJOR ACTIVITIES

This section of the paper describes the three major activities accomplished during the implementation of the statewide pavement management system: database development, selection of a pavement condition data collection service, and selection of pavement management system software. There were many other activities involved in accomplishing the implementation of the system, but these three are the technical tasks requiring the greatest level of resources. Further, a great deal of time was devoted to mitigating or dealing directly with nontechnical institutional issues.

5.1 Database development

The IPMP basic design required the development of a geographic information system (GIS) database supported by dynamic segmentation. Dynamic segmentation provides for more flexible data management without requiring the duplication of network geometry or data. Dynamic segmentation also allows for the use of multiple linear referencing methods, while at the same time accommodating statically segmented data sets. GIS for data analysis allows flexibility in displaying and integrating all of the pavement data.

Figure 1 is intended to illustrate our use of dynamic segmentation. In Figure 1, the first horizontal bar represents a longitudinal view of the highway. Along the highway, several types of data are collected that are germane to pavement management. The second horizontal bar represents the same highway, but the highway is divided into pavement management sections. The third horizontal bar represents the pavement test sections. Across each test section the pavement condition (e.g., surface distress and roughness) is measured. On Iowa's primary system, the beginning and end points of pavement management sections are referenced by kilometer points along a route. The kilometer point referencing method (originally a mile point system) is the legacy system used to locate pavement management sections and many other features along a route. Beginning and ending locations of pavement test sections are located using latitude and longitude using differential-corrected global positioning system (DGPS) measurements. Dynamic segmentation matches the location of the test section with respect to the pavement management section by knowing the distance along the route for each reference point regardless of the referencing method used (e.g., using either latitude and longitude, kilometer points, or another system). The same is true for information in the highway inventory (e.g., information on the pavement cross section, construction history, and traffic volumes), the bottom most horizontal bar. Thus, each database may reside in its own native location referencing method and is related to the other databases through dynamic segmentation.

In our case, we are interested in aggregating pavement data into pavement management sections. As an example, the pavement condition measurements in each of the test sections must be aggregated to one representative measure of the entire pavement management section. A simple method of aggregation is to take the average distress measurements of the test sections covered by the pavement management section and set the average distress level as being representative of the entire pavement management section. On the other hand, we may wish to develop more sophisticated logical rules for aggregating condition. For example, with safety-related pavement condition measurements (e.g., rut depth) the aggregation rule may be to make the worst pavement condition measurement of any test section representative of the entire pavement management section. Using dynamic segmentation, logic rules are used to aggregate or disaggregate pavement data into measurements that are most representative of the entire pavement management section.

Figure 1. Dynamic Segmentation

Data for the IPMP originate from three different sources shown in Figure 1. The principal source of data is the Iowa DOT's highway inventory known as the "Base Records." These inventory data are based on a kilometer point (km.pt) referencing system. Data characterizing the inventory sections (attribute data) are stored using variable-length static segments. Each record is identified by a route name, a begin km.pt, and length. Base record sections have variable length (0.01 to 1 mile).

Local agencies (cities and counties) and the Iowa DOT provided history data for all of the FAE highway network for each pavement management section. The Iowa DOT and some local agencies already had in place a pavement management system, and they have already defined pavement management sections. Local agencies with a pavement segmentation scheme in place were asked to provide us with their segmentation. Local agencies without an existing pavement management system were provided sectioning guidelines and asked to develop management sections. Section beginning and end points from local governments are also stored using variable-length static segmentation, but the information is referenced using literal description (from point X to point Y). Each record contains the surface type, maintenance and rehabilitation history, and any available pavement construction cost information.

Pavement condition data are automatically collected and stored using fixed-length static segmentation and referenced by geographic latitude and longitude coordinates recorded using DGPS. Test sections are constant in length, each covering 0.1 km. The types of distress information collected are discussed later in this paper.

Following is a description of the basic steps used to develop the database.

  • Developing the graphic linear network: A base map of the entire state network of roadways (all FAE roads operated by all levels of agencies) was created using data from the Iowa DOT. The Iowa DOT created the data through a cartography attribution process.
  • Storing and managing data: To facilitate data analysis, the IPMP data are stored and managed according to regional planning affiliation (RPA) limits. Individual data sets within each RPA are aggregated by county and city. Also, the state system is maintained in a separate data set. The use of dynamic segmentation provides the capability to manage these different data sets and will enable data analysis on a statewide basis.
  • Data maintenance: The attribute data associated with the graphic linear network must be updated whenever highway alignments change. Also, pavement management section limits might change from year to year. Therefore, maintenance procedures are established to update the data and maintain the integrity of the database.
  • Data access: The IPMP relational GIS database is quite complex. An inexperienced user may find it difficult to perform simple data analysis. Further, some users may want access to the data for use in their own native analysis systems, ranging from paper records to sophisticated pavement management software systems. Thus at least three forms of data access are being made available: 1) paper copies of IPMP data for individual jurisdictions or RPAs, 2) electronic copies of IPMP data for individual jurisdictions or RPAs in a variety of common formats (e.g., text files, Lotus or Excel files, etc.), and 3) electronic copies of IPMP data for individual jurisdictions or RPAs formatted for GIS software for individual local jurisdictions or RPAs/MPOs. A fourth level of access will be provided through an inexpensive desktop GIS which will function as an executive information system.
  • Management systems data integration: One of the overarching objectives of the IPMP was to integrate pavement management information with other transportation system management systems. The design of the IPMP database allows for data sharing among different management systems databases and provides a compatible framework to be used for other database developments. In fact, critical elements of the IPMP are exported into an integrated management system that combines critical data from several systems to support resource allocation decisions across management areas. The integration of data is shown graphically in Figure 2.

Figure 2. Integrated Management Systems Databases

5.2 Pavement condition data collection

One of the key objectives of the IPMP is to create an unbiased information system to support pavement management decision making at all levels of government. In order to minimize the possibility of bias, the IPMP task force recommended the use of automated pavement condition data collection technology. The IPMP data collection committee was charged with this task. The committee, which consists of city, county, and Iowa DOT representatives, focused on the following issues:

  • Identifying categories of pavement types based on paving materials, where pavements in each category perform similarly, or similarly enough, to be included in the same pavement performance models. The categorization must be amply broad to allow its application across all three levels of jurisdictions (i.e., state, city, and county roadways).
  • The requirements for pavement condition data elements (distress types) necessary to support network-level decision making at each level of jurisdiction.
  • The requirements for pavement condition data elements (distress types) necessary to support project-level decision making at each level of jurisdiction.
  • The coverage of the data collection (e.g., could data be collected for a representative sample, or was it necessary to have 100 percent coverage).
  • The necessary frequency of data collection (e.g., collect pavement condition data every year or less frequently).
  • The feasibility and effectiveness of using automated data collection equipment to collect condition data to support the data requirements for pavement management at each level of jurisdiction.

The data collection committee developed recommendations for each of the above issues (the results are discussed in the next subsections) and proceeded to develop an evaluation and selection process for a pavement condition data collection methodology. Two options were considered. the first was to purchase a service agreement with a vendor for pavement condition data collection, reduction, and delivery. The other was to purchase equipment and have Iowa DOT personnel operate the equipment.

Because automated pavement condition measurement technology is still evolving, the recommendation was made to purchase a service agreement for data collection. After an evaluation, including field trials of the available technology, the committee recommended contracting for services with Roadware Corporation, Paris, Ontario, Canada, to collect the condition data for the FAE highways in Iowa (Smadi et al, 1997).

5.2.1 Coverage and frequency

The recommended data collection coverage and frequency are the result of a compromise among the agencies' needs, cost, and the vendor capabilities and limitations. The recommendations were made for the pavement condition data to be collected in one direction on two-lane highways and two directions on multi-lane highways. Test sections will be 0.1 km for both rural and urban areas. Pavement condition data are collected for 100 percent of the test section. Using dynamic segmentation, the distresses from each 0.1 km test section will be summarized for each pavement management section. The pavement condition data are collected over a two-year cycle, resulting in covering half of the network every year. Test section locations are identified by latitude and longitude using DGPS.

5.2.2 Pavement condition data collected

The committee's recommendations are based on providing enough information to support network-level pavement management for asphalt concrete and portland cement concrete surfaced pavements. Table 1 describes each distress and its severity and extent. Both ride and rutting data will be automatically collected for all test sections.

Table 1: Distresses collected by pavement type

PCC Pavements Description
Joints - D-Cracking

Extent:

Severity:

Number of joints with D-cracking

SHRP definition for Moderate and High

Joints - Spalling

Extent:

Severity:

Number of spalled joints

SHRP definition for Moderate and High

Transverse Cracking

Extent:

Severity:

Number of transverse cracks

SHRP definition for Moderate and High

Patching

Extent:

Severity:

Area and number of patches

Distress or no distress

ACC Pavements Description
Transverse Cracking

Extent:

Severity:

Number of transverse cracks

SHRP definition of Low, Moderate, and High

 

Longitudinal Cracking

Extent:

Severity:

Length of longitudinal cracks

Sharp definition for Moderate and High

 

Determine extent and severity for wheel path and non-wheel path

 

 

Block Cracking

Extent:

Severity:

Area of block cracking

SHRP definition of Moderate and High

 

Change in the definition of the piece size from 0.1 to 1.0 sq. meter

 

Alligator Cracking

Extent:

Severity:

Area of alligator cracking

SHRP definition for Moderate and High

Potholes

Extent:

Severity:

Number of potholes

Distress or no distress

Patching

Extent:

Severity:

Area and number of patches

Distress or no distress

5.3 Pavement management software

At the time of this writing, pavement management software has not yet been selected. The IPMP pavement management software committee is still in the process of evaluating different software options to meet the needs of both local and regional governmental agencies using the pavement management software.

The IPMP task force had approved a set of recommendations for the PMS software. The recommendations deal with three different aspects of PMS. The first aspect is related to the analysis level, project or network. The second deals with the issue of performance forecasting of the pavement condition in the future. Finally, the last aspect is the resource allocation decision-making tool (prioritization or optimization).

The task force's performance requirements have been divided between those for the network-level analysis and those for the project-level analysis. The software systems (e.g., one for network- and one for project-level analysis) or system must satisfy the requirements identified by the task force and listed below.

5.3.1 Network-level software

The principal focus of the IPMP has always been to support network-level pavement management, although the same data can be used to conduct project-level pavement management. The IPMP will provide software and training to allow city, county, and regional staff members to perform network-level pavement management within their own jurisdictions. The selected software must support the following functions:

Conduct network-level pavement management analysis that produces the following:

  • Resource allocation strategies for cost-effective pavement management (maintenance, rehabilitation, or reconstruction) decisions.
  • Provide current and future pavement network conditions based on:
    • Funding levels
    • Performance constraints
    • Other user-defined variables

Utilize mathematical programming techniques for the optimal allocation of resources based on user-defined optimization and analysis parameters.

Initially utilize probabilistic performance prediction techniques for the condition of the pavement network.

5.3.2 Project-level software

Similarly, the software selected must support project-level analysis by cities, counties, and regional governments. The IPMP will also make software and training available to help governments conduct project-level analyses. The software selected must support one or more of the following functions:

  • Single-year prioritization
  • Multi-year prioritization with performance trends
  • Multi-year optimization (benefit cost, incremental benefit cost, or marginal cost analysis)
  • Provide for short-term analysis (three to five years)
  • Reporting (tabular and/or graphical), querying, and viewing capabilities

6. DISCUSSION AND CONCLUSIONS

This paper deals with the technical issues related to the development of a statewide pavement management database. Although the emphasis in this paper is on technical issues, nontechnical institutional issues are of equal or greater importance than technical issues and consumed must of the project's resources, particularly in the early stages of the system's design.

The purpose of the Iowa Pavement Management Program is to prepare the pavement database and pavement management software necessary to support the conduct of pavement management by Iowa's roadway operating agencies (cities, counties, and the state) and regional governments. Pavement management and pavement resource allocation will continue to be conducted under the auspices of the highway operating organizations, while supporting activities are conducted centrally through a standard technical process. By centralizing the support mechanisms, local and regional governments have access to far greater pavement management and database expertise than they would otherwise, as well as to higher quality pavement condition data and superior pavement management decision-making tools. Centralizing pavement management support services and distributing pavement management decision making to each of the highway operating agencies efficiently divides responsibilities while encouraging the statewide use of pavement management systems.

The platform for the IPMP is a statewide GIS database with dynamic segmentation capabilities. Initially it was recognized that such a sophisticated database was not required to support the data needs of pavement management. However, dynamic segmentation and GIS provide a more flexible environment that allows the integration of pavement management data with other highway attribute data and activity data (e.g., accident data or maintenance work data). Using the pavement management database as a platform, other integrated transportation databases have been developed, providing added returns on the investment made in a GIS database with dynamic segmentation capabilities.

7. REFERENCES

T.H. Maze and Omar G. Smadi, "Taxonomy of Institutional Barriers to the Implementation of Pavement Management Systems," Proceedings of the Seventh Maintenance Management Conference, National Academy Press, Washington, D.C., 1995.

Omar G. Smadi, Brian McWaters, Kevin Jones, Robert Gumbert, Randall Kraull, and T. H. Maze. "Selection of an Automated Pavement Distress Technology: Iowa Case Study." Presented at the Seventy-Sixth Annual Transportation Research Board Meeting, Washington, D.C., January 1997.

CTRE is an Iowa State University center, administered by the Institute for Transportation.

Address: 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664

Phone: 515-294-8103
FAX: 515-294-0467

Website: www.ctre.iastate.edu/

Iowa State University--Becoming the Best