Featuring developments in Federal highway policies, programs, and research and technology.
|This magazine is an archived publication and may contain dated technical, contact, and link information.|
|Federal Highway Administration > Publications > Public Roads > Vol. 62· No. 1 > Pavement Management Systems — Past, Present, and Future, Public Roads: 80 Years Old, But the Best Is Yet to Come|
Pavement Management Systems — Past, Present, and Future, Public Roads: 80 Years Old, But the Best Is Yet to Come
by Fred Finn
This article was adapted from a presentation by Fred Finn at the National Workshop on Pavement Management in New Orleans, La., July 20, 1997.
It is hard to say when the idea of systematically managing pavement networks first started.
Probably some unknown engineer in some unknown state decided he (it was probably a "he" because this took place 30 or 40 years ago) needed to measure the condition of the pavements in his state to better prioritize his maintenance activities. Using note cards, he probably identified the location, limits, and condition of the roads and who lived along the route or what businesses were effected. Maybe it was a Roman engineer charged with maintaining the Appian Way some 2,000 years ago. We will never know for sure and probably don't really need to know.
We do need to know and agree on the goals for pavement management and how those goals can be achieved and how to convince top management that the implementation of such systems is in the best interest of the public to which they are accountable.
The first national workshops on pavement management, held in 1980 in Phoenix, Ariz., and Charlotte, N.C., were a beginning to an understanding of these goals. These two workshops brought together a total of 172 engineers and scientists who were interested in learning more about pavement management and pavement management systems (PMS). The bottom line or goal of the workshops was to "formulate and set priorities for a national program of short- and long-term activities to be pursued by FHWA [Federal Highway Administration] and state agencies toward better management of pavements."
In 1980, only five states - Arizona, California, Idaho, Utah, and Washington - were reported to be in various stages of development of systematic procedures for managing pavement networks on a project-by-project basis. Utah, under the leadership of Dale Peterson, was probably the first state to publish information concerning the possibilities of pavement management in a 1977 paper titled "Good Roads Cost Less." Expressing a pay-me-now or pay-me-later perspective regarding the best time to perform rehabilitation on main-line pavements, the paper established that timely maintenance would save money over the long term.
Now, all 50 states, the District of Columbia, and Puerto Rico have some form of pavement management programs in place or in development.
In many respects, the beginning of pavement management systems started with the AASHO Road Test from 1956 to 1960. The road test staff determined that it would be necessary to evaluate the performance of pavements in a way that would be independent of pavement type and that could have universal application for describing a pavement's condition. The method developed and used at the road test is based on a pavement's present serviceability (riding comfort).
As I have traveled around the country since completion of the road test, I continue to find some misunderstanding about how and why this concept was initially developed. The working hypothesis for the development of present serviceability was based on the belief that riding comfort, along with safety, were the primary performance objectives to be associated with pavements.
The problems were how to estimate riding comfort without having to actually ride over every section on the road test every two weeks and how to translate this concept for use by state agencies.
The solution was pretty simple in concept, as described by Carey and Irick in "The Pavement Serviceability-Performance Concepts," which was published in Highway Research Bulletin 250 in 1960. First, assemble a group of people to ride over selected sections and get their respective opinions on the ride quality of the section. Second, obtain physical measurements of the condition for each of the sections, and third, correlate the subjective responses to the physical conditions.
At the road test, the subjective estimates were obtained by a panel of 12, mostly engineers, riding over selected pavements that were judged to represent a wide range of conditions. Ratings were made on a continuous scale from 1 to 5 and were described as very poor, poor, fair, good, and very good. The panel was cautioned against being influenced by the appearance of the pavement; they were to focus only on the riding comfort or the quality of the ride. This subjective rating is called the Present Serviceability Rating (PSR). After rating each section, the panel was asked to indicate if the pavement was considered acceptable or unacceptable. One guideline for this response was in the form of a question: "How would you feel about driving over the rated section for a long period of time or a long distance - would it be acceptable or unacceptable?"
The next step was to make the necessary physical measurements on the rated pavements. The usual forms of distress, such as rutting, faulting, cracking, patching, and raveling, were measured. Because serviceability (riding comfort) was assumed to be a function of pavement roughness, this property was also measured using the AASHO-developed Clhoe (Carey, Leathers, Huckins, and others) profilometer. This device was designed to measure roughness as a function of slope variance in the wheel path, using two small wheels closely spaced. Finally, correlations were made between the average panel ratings (PSRs) and the physical measurements on pavements in Illinois, Indiana, and Minnesota. A method was found to correlate objective measurements with the average of the subjective responses. This correlation is called the Present Serviceability Index (PSI). Carey and Irick indicated that about 95 percent of the information about the serviceability of a pavement is contributed by the roughness or surface profile. Other factors, such as cracking, rutting, and patching, were considered statistically significant but made only minor improvements in the estimates. It is no wonder that many agencies have relied solely on road profile or measures of roughness to estimate serviceability. As might be expected, there were criticisms about the way in which the subjective ratings were obtained but not about the concept. The concept of riding comfort, as reflected in measurements of roughness or profile, to estimate performance remains the keystone of most PMS although not always the primary triggering factor to initiate maintenance or rehabilitation.
The measurement of roughness and its relationship to riding comfort has been the subject of millions of dollars of research since the completion of the road test. Much of this research has been directed to developing better ways to measure roughness or some response to roughness that can provide more reliable estimates of riding comfort. This effort will no doubt continue.
For the present, the use of the International Roughness Index (IRI) is the principal method used to measure roughness and to relate it to riding comfort. In a survey by FHWA in late 1996, 47 of the 52 respondents said they use IRI to measure roughness as a surrogate for riding comfort. There may be better ways to estimate ride quality or comfort, but consistency in the method of measurement is important to develop the prediction models needed for a comprehensive pavement management system, and IRI meets that criteria for the present. In 1981, there was considerable controversy over the use of the word "system" as it applies to pavement management. Amazing as it may sound, much time was spent in discussion and argument about which term to use: pavement management or pavement management systems. The dictionary defines system as: "a set of facts, principles, rules, etc. classified and arranged in a regular, orderly form so as to show a logical plan linking the various parts." It always seemed to me that this definition helps to define the specific aspect of management we want to achieve with this process.
In 1970, FHWA and the Highway Research Board (now the Transportation Research Board) sponsored a workshop, organized at the University of Texas, to discuss structural design of asphalt concrete pavement systems. The workshop participants, probably for the first time at a national level, discussed the pros and cons of systems engineering and its potential application to pavement design and a hint of possibilities for pavement management.
At this meeting, Dr. Karl Pister, a professor of civil engineering (and later dean of the College of Engineering at the University of California at Berkeley and subsequently chancellor of the University of California at Santa Cruz), described the potential benefits of systems engineering. Pister presented an approach for estimating "optimality" in the decision-making process as part of a management system. He even provided general mathematical solutions to achieve such optimality for a pavement management system. Pister did not invent systems engineering, he simply (maybe not so simply) pointed out that pavement design and pavement management were very complicated problems and that one way to handle complicated problems is through the use of systems engineering.
Also in 1970, Haas and Hutchinson presented a paper to the Australian Road Research Board on "A Management System for Highway Pavements."
In 1974, the Washington State Department of Transportation (DOT) implemented a project-level pavement management system with models and timing recommendations for various types of maintenance and rehabilitation actions.
In 1977, Ralph Haas and Ron Hudson published the first edition of their textbook, Pavement Management Systems.
In 1979, NCHRP (National Cooperative Highway Research Program) Report 215 titled Pavement Management System Development by Hudson, Haas, and Pedigo was issued and helped to establish a "how to" guide for development of PMS.
From 1968 to 1980, there were a number of engineers and scientists who were convinced that a pavement management system was a good idea and were willing to push ahead. At the risk of missing some, I want to mention a few who carried both the technical and managerial banners:
After 1980, the list becomes too long to mention.
Some of the first management systems consisted primarily of a database, a condition index, and a ranking system that was used to develop a prioritized list of projects following a "worst first" philosophy. The ranking systems subjectively weighed factors, such as roughness, cracking of various kinds, raveling, rutting, and spalling, and produced a combined score or index. Some condition surveys included as many as 15 categories to be evaluated and recorded. Indices of this kind are still used by many agencies as a way to summarize pavement conditions within a specific network. The U.S. Army Corps of Engineers developed a somewhat more rational way of calculating an index, but in the final analysis, it too is based largely on engineering judgment.
The participants in the 1980 national workshops in Phoenix and Charlotte attempted to evaluate the possibilities for the development and implementation of PMS. Some of the significant products from these workshops were summarized in the Proceedings published by FHWA in June 1981.
One of the products included a definition for pavement management system: "a system which involves the identification of optimum strategies at various management levels and maintains pavements at an adequate level of serviceability. These include, but are not limited to, systematic procedures for scheduling maintenance and rehabilitation activities based on optimization of benefits and minimization of costs." A 13-step plan was proposed for the development and implementation of a pavement management system:
A differentiation was made between project management and network-level management by the participants at the 1980 workshops. This was significant because in 1980, many were thinking only in terms of individual projects rather than the network.
The best or optimal recommendation at the project level may not be the optimal solution for the network. The easiest illustration of this relationship can be posed by the question: Do I spend a major share of funds available on one or two projects to achieve acceptable performance over an extended period of time, or do I distribute the funds over several projects with reduced expectations of the service life but an overall raising of the serviceability of the network? The answer can only be addressed at the network level, and depending on a variety of factors or considerations, such as user costs, the final decision could go either way.
While a certain amount of enthusiasm for the concept of PMS was produced at the 1980 workshops, there were also several "buts," as there always are when exploring new concepts.
Some people were concerned about the ability to predict the long-term performance of pavements for the types of distress that can occur and that could trigger maintenance or rehabilitation. Most agreed that we needed to apply engineering judgment and institutional experience about which specific distresses to include because many can be handled by maintenance management systems that, by 1980, were being developed by many state and local agencies. Three to five distress types and a measure of roughness were sufficient for most states. In my view, the major problems that contributed to skepticism about the viability of PMS in 1980 and beyond were:
But the most significant factor was very simple. Many people simply did not believe that such a system would work. While the consensus held the perspective that the potential benefits could provide a powerful and more objective way to program maintenance and rehabilitation activities, some said it would not be able to reliably predict performance, it would not be able to recommend the best treatment at the best time, and it would not result in any improvements to the highway network.
We must remember that, for the most part, pavement engineers are trained as civil engineers. In addition to engineering experience, PMS operations require a knowledge of statistics, modeling, economics, theories of optimization (operations research), computer science (sophisticated programming), and database management. It is natural to tend to resist multidisciplinary systems that are outside our realm of training, experience, and understanding.
But that was then and this is now.
There have been three International Conferences on Pavement Management, and a fourth is planned in South Africa in 1998. The Transportation Research Board established a Pavement Management Section with committees on Pavement Management Systems and Pavement Rehabilitation. Literally hundreds of papers have been presented and published in Records of the Board. The American Association of State Highway and Transportation Officials (AASHTO) issued "Guidelines for Pavement Management Systems." The World Bank has its own pavement management system for specific use in developing countries. The American Society of Civil Engineers has published a number of articles on the subject in the Civil Engineer magazine. The National Highway Institute continues to sponsor a large number of in-service training courses to help state highway agencies better understand various aspects of PMS.
In addition to acting as a national facilitator for the coordination and encouragement of state DOTs, FHWA has played a key role in supporting the development and implementation of PMS through their support of the International Conferences and through incentives provided in the federal legislation, including the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA). In ISTEA, FHWA identified pavement condition indicators that are incorporated in PMS as a basis for evaluating requests for interstate maintenance funds. And FHWA policy said that "Each State [was] to have an operational pavement management system for principal arterials in place by January 13, 1993." While many of those requirements have since been eliminated, the recognition of the value of PMS, on a national level is clearly indicated by its inclusion in policy statements of FHWA and in similar statements by some State DOTs, regarding the management systems for counties and cities.
In 1996, FHWA and AASHTO sponsored a seminar on Asset Management that was attended by representatives of industry, academia, and decision-makers from federal and state transportation agencies. In my opinion, this seminar should help to bring together a better understanding of the similarities that exist between asset management as viewed by private industry and governmental agencies - especially agencies charged with management of our infrastructure system. A number of statements from that seminar are worth repeating:
The FHWA data bank indicates that each of the 50 states, the District of Columbia, and Puerto Rico have some kind of a pavement management system. Forty-two states reported that their systems included a method of prioritization, and 20 states included a provision for optimization for purposes of budget planning and project programming.
While 47 of the 52 agencies use IRI to measure ride, they use 10 different types of equipment to measure IRI, according to a report prepared for FHWA by the Texas Research and Development Foundation in 1994. In addition to the variation in equipment, the method of measurement is not consistent; some measure the right-side wheel path only, some the left-side only, some report only the worst, and some average values from both wheel paths.
Among the 19 states that provided information about criteria for determining quality levels, the beginning of the unacceptable (poor) level for IRI measurements ranged from 450 inches per mile (7.09 meters) to 45 inches per mile (0.71 meters). It would seem clear that the major causes of these differences are related to the different equipment used and the different methods of measurement. Such differences in equipment, methods of measurement, and reporting make it difficult to compare information and to share experience and to work together for improvements.
According to a 1996 survey by FHWA, which included information from the 52 agencies, the dominant forms of distress being measured and included in respective PMS databases are rutting, faulting, and cracking. Surface friction information is measured and stored by 39 agencies. Five states measure deflection at the network level, and nine have deflection measurements under development. Clearly, the implementation of PMS databases is generating a large amount of empirical and useful information on the condition of pavements. This information is needed for prioritization, optimization, and the development of prediction models.
The TDRF report, which included information from 48 states, makes it clear that there is a divergence in the way in which rut depth and faulting are measured and reported. For example, 86 percent of the reporting states indicated they have the capability to measure rut depth using automated equipment, while 7 percent measure manually, and 7 percent visually. In the states that use equipment to measure rut depth, they use six different types of equipment with the number of sensors varying from three to 37. In quantifying the levels of rut depth, the boundary between "fair" and "poor," which can be translated as barely acceptable to unacceptable, ranged from 1 in (25 mm) to 0.5 in (12.5 mm). Differences also were reported in the method of sampling. Methods included mile-by-mile, construction section, homogeneous section, and a category called "others." The measurements may be representative of the inner wheel path, the outer wheel path, or some combination of both wheel paths. Five different methods of equipment calibration were reported, ranging from accepting the vendor's calibration to no calibration at all. Incidentally, of the 48 states responding, 25 states (52 percent) indicated rutting was considered a problem in their state. Whether rutting is a problem or not may depend on the levels associated with the fair and poor categories for which there does not appear to be a consensus.
Faulting data is collected by 29 states. Five states have no rigid pavements on the interstate system, eliminating the need for such measurements. Eight states used automated equipment from three different vendors. A range of "faultmeters" were also reported as being used, and seven states make visual estimates. Unacceptable levels of faulting ranged from greater than 0.25 in (6 mm) to 0.75 in (19 mm). As in the case of rut depth, the methods used to sample a section varied, and the manner used to summarize the data into information also varied.
As always, there are pros and cons regarding the desirability of these protocols at this stage in the development of PMS and the potential impact on the utility of historical database information. However, I believe there is some consensus among users that it would be beneficial if all of the agencies were measuring, summarizing, and reporting comparable information. It would facilitate communications, and it would add a synergistic aspect to data collection and analysis for which the resulting information would be more valuable than the sum of the individual parts.
"There should be a desire for change by skillful professionals who can make a difference in the way decisions are made and which will result in reduced costs and improved performance," said Dr. Tom Larson, a former FHWA administrator, as part of his remarks at the 1985 International Conference on Pavement Management. He pointed out that "without dollars there will be no need for pavement management and without good management there will not be enough dollars to go around."
It seems clear that pavement management and pavement management systems have been recognized as a legitimate part of highway engineering and as a viable tool for the decision-makers in managing the rather considerable resources entrusted to them in the form of pavements.
Technology has advanced rapidly in the past 10 years, and it is possible to develop prediction models - methods to optimize choices between competing alternatives and to develop multiyear prioritization, remaining life, life-cycle costs, and feedback information - if we chose to do so. In 1970, when I first became interested in PMS, we did not have such tools, much less access to computer hardware, to perform all of the necessary calculations. Now it can be done on the computer on your desk. At times, you may be overwhelmed with the possibilities and the choices, but you do have choices.
I am not suggesting that development and implementation is a walk in the park; it is not. Unresolved issues include: the reliability and credibility of probabilistic prediction models and how uncertainty can be accommodated by a PMS, uncertainty regarding cost information due to variations within a state and over time, consideration given to user costs, credibility of the maintenance and rehabilitation (M&R) recommendations, and the adequacy of the performance factors and consistency of measurements.
The technology required to develop and implement a pavement management system is available. The main problems stem from institutional issues and the lack of a solid, long-term commitment to implement a pavement management system as part of an overall infrastructure management that includes bridges. While some of the commitment must come from managers who control resources, some must come also from engineers involved in development and implementation. This commitment should be in the form of an optimistic and positive attitude that will overcome any negativity within your organization. If you are not positive and optimistic about your ability to operate a pavement management system, it is doomed to failure.
Full use is also important. Some states have developed very comprehensive systems, including both budget requirements and M&R programming, but they have chosen to use only the budget preparation part of the system. A major benefit of PMS is to help make decisions concerning the what, where, and when of M&R actions. Without this element, the optimization aspects of the system cannot be achieved.
PMS cannot make decisions; only the decision-maker can do that. However, the question is whether the decision-makers are effectively using the best tools to assist themselves in making the best decisions. I hate to close with a cliché, but I will. Surely, if we can put a man on the moon and a robot on Mars, we can design, develop, and implement a workable and productive pavement management system.
Fred Finn retired in 1993. He was licensed as a professional engineer in six states - Arizona, California, Idaho, Nevada, Oregon, and Washington - and he retains his license in Arizona. He has 44 years of experience through academia, consulting, and the military in all aspects of pavement engineering, including research, design, construction, and maintenance. Since 1970, he has been involved in the application of systems engineering to pavement design and management systems. He has helped develop pavement management systems for several states and has served as a consultant to the World Bank for the development of a highway management system for pavements, structures, and roadside elements in Saudi Arabia. He is a member of the Transportation Research Board, a Fellow in the American Society of Civil Engineers, and past president and honorary member of the Association of Asphalt Paving Technologists. In 1993, he was elected to the National Academy of Engineering. He has a bachelor's degree in civil engineering from the University of New Mexico and a master's degree in civil engineering from the University of California at Berkeley. He now lives in Monticello, Ill., where he is very active in community affairs.
Page Owner: Office of Corporate Research, Technology, and Innovation Management
Scheduled Update: Archive - No Update
Technical Issues: TFHRC.WebMaster@dot.gov