Conference Paper

Sensitivity Analysis of Performance Metrics to Different Parameters in Pavement Management System

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

A pavement management system is a useful tool for departments of transportation to address the problems of limited budget and aging infrastructures. Previous research focuses mainly on the budget allocation process, trying to improve the optimization algorithms and consider the uncertainty of predictions for pavement deterioration. In any given pavement management system, there are usually many parameters. However, analysis has not been performed to determine the influence of different parameters on the pavement network performance. In this paper, the sensitivity of performance metrics to different parameters is explored based on the interstate pavement network in the U.S. state of Virginia by a probabilistic allocation network model developed at MIT. A statistical method is applied to conduct the sensitivity analysis. The sensitivity of performance metrics to different parameters is decided by p values, and the relative significance of different parameters is compared and ordered by z-score statistics.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... To overcome these challenges, multiple policies and acts were legislated in the United States, such as the Moving Ahead for Progress in the 21st Century (MAP-21) and Fixing America's Surface Transportation (FAST) acts (Zimmerman 2017). These two acts require Departments of Transportation (DOTs) to implement Pavement Management Systems (PMSs) in order to improve pavement performance (Guo et al. 2018). A PMS provides decision-makers with tools to help them evaluate current condition, predict future condition, and determine optimum maintenance and rehabilitation (M&R) budget allocation strategies. ...
Article
Full-text available
Treatment records are among the most frequently underreported data items in pavement management systems (PMS), which negatively affects various PMS analysis tools, such as pavement performance and deterioration models. Disregarding unreported treatments may lead to inaccurate pavement age and condition estimates, resulting in erroneous and non-optimal maintenance and rehabilitation decisions. Nevertheless, the unreported and frequently missing pavement treatment data received limited attention. Hence, this paper contributes to the body of knowledge by introducing a methodology for detecting unreported treatment actions and their occurrence probabilities over pavement age using a machine learning classification algorithm. Logistic regression models were developed using historical pavement condition data and validated on two levels: (1) split validation and (2) manual validation using video logs of pavement condition before and after treatment application. The results show that the developed models can detect unreported pavement treatments with accuracy, precision, and F1 scores ranging from 89% to 96%, 82% to 91%, and 70% to 85%, respectively. The presented methodology and developed models will help highway agencies identify unreported and missing pavement treatments, contributing to more cost-effective maintenance and rehabilitation decisions.
... In most PMSs, numerous parameters have the potential to affect optimum maintenance strategies. Yet parametric analysis has not been performed to test the effects of changing these parameters (Guo et al. 2018). The result is insufficient information on the consequences of changing pavement management parameters and maintenance policies on optimization outcomes. ...
Article
Pavement maintenance policies are being adapted for different road classes in the decision-making process of the pavement management system (PMS). Optimization analysis is a promising approach to managing maintenance strategies considering pavement performance targets and budget constraints. One approach to increasing the effectiveness of optimization analysis is to study the influences of changing key parameters in PMSs on optimum maintenance strategies. In this paper, a multiyear optimization analysis is conducted while managing low-volume roads (LVRs) in Colorado at the network level. First, the regional pavement maintenance practice is defined for five engineering regions managed by the Colorado DOT. Then a sensitivity analysis is used to show the effects of changing maintenance policies of thin overlays, size of road network, pavement performance targets, and initial pavement conditions on the optimum solutions. The results of this study provide better understanding of the importance of certain criteria when justifying maintenance funding needs and maintenance policies. Engineers can consider the factors investigated in this study to secure appropriate maintenance funding for LVRs from state legislatures.
Article
Full-text available
Management of large transportation infrastructure, such as pavements, requires sound management tools to guide activities in an optimal fashion. Several pavement management systems (PMS) are in use by many transportation agencies. There are two types of management systems. There are those systems that are driven by optimization models and those that rely on expert (knowledge) systems. As transportation networks become larger and larger, it becomes imperative that some form of optimization modeling is needed. The majority of such optimization models are of the type where a single objective is optimized. In reality however, transportation officials are faced with multiple objectives or criteria that need to be traded off each other before a final decision is reached. In this paper, the author proposes a framework for developing a stochastic, multi-objective PMS. The paper discusses the multi-objective framework and highlights the departure from its predecessor; stochastic single-objective PMS. The paper shows that this proposed approach would result in several non-dominated (Pareto-optimal) policies. These are PMS policies whose attributes make them competitive. The advantage of this approach is to allow decision makers to have an important leverage in the final choice of PMS policy. The paper also outlines added advantages to this approach. These include possibility of generating dual equivalence (e.g., distress-based and roughness-based) policies that have similar network consequences. This feature is shown to have potentials for developing maintenance trigger levels in both scales. The feature could also enable agencies to develop distress-roughness equivalence tables where equivalent distress/roughness levels that would trigger same treatment actions may be developed.
Article
Full-text available
We present an application of systematic probing for selecting optimal maintenance, repair, and reconstruction MR&R policies for systems of infrastructure facilities under model uncertainty. We use an open-loop feedback control approach, where the model parameters are updated sequentially after every inspection round. The use of systematic probing improves the convergence of the model parameters by ensuring that all permissible actions are applied to every condition state. The results of the parametric analyses demonstrate that the MR&R policies converge earlier when systematic probing is used. However, the savings in the expected total costs as a result of probing are minor, and are only realized when the optimal probing fractions are used. On the other hand, the additional costs incurred when the wrong probing fractions are used are significant. The major conclusion from this work is that state-of-the-art adaptive infra-structure management systems, that do not use probing, provide sufficiently close to optimal policies.
Article
Priority analysis is a multi-criteria process that determines the best ranking list of candidate sections for maintenance based on several factors. In this paper, two methods for priority ranking of road maintenance, viz. (a) ranking based on subjective rating and (b) ranking based on economic indicator, are evaluated. The subjective ranking was done using maintenance priority index which is a function of road condition index, traffic volume factor, special factor and drainage factor. The second ranking method was based on economic indicator in which NPV/Cost ratio was calculated for each pavement section using the HDM-4 software.
Article
A pavement management system was developed for the State of Arizona to produce optimal maintenance policies for each mile of the 7,400-mile network of highways. The heart of the system is a mathematical model which captures the dynamic and probabilistic aspects of pavement maintenance. The model integrates management policy decisions, budgetary policies, environmental factors, and engineering decisions. During the first year of implementation, fiscal year 1980–81, the system saved 14 million dollars (almost 1/3 of Arizona's preservation budget). The forecast savings over the next four years is 101 million dollars.
Article
A global network optimization model has been developed for generating a pavement repair plan using expected performance of pavement repair works. The expected performance of pavement repair works is represented by the expected age (service life) associated with potential repair actions. The expected age defines the anticipated pavement condition improvement obtained from applying a particular repair action. The expected age for each repair action is usually known from either experience or assumed as part of a design procedure. A constrained linear optimization model is formulated with its main objective of optimizing the expected pavement condition improvement. Pavement condition improvement is defined as the age gain in year lane kilometer or average age in years extended to a pavement network as a result of applying potential repair actions. The global linear model is subjected to a single budget constraint enforcing the total budget available for the entire network and limitation constraints placing lower and upper limits on the repair variables. The optimum repair plan provides a macroscopic solution as the repair variables represent pavement proportions that should be treated by the corresponding repair actions. The pavement network is divided into a number of systems with similar pavement structures and loading conditions. Presented sample results have indicated that global network optimization of the pavement management problem may not result in a rational budget allocation among deployed pavement systems. This problem can be solved by enforcing system improvement requirement constraints.
Article
This paper presents a need-based methodology to prioritize and select highway projects for improvement. This approach is based on developing a multiattribute need function that quantifies relative concerns of a highway agency and the traveling public about various physical and operational deficiencies on different highway segments. It is an effective alternative to the traditional cost-benefit analysis. While the cost-benefit analysis may be useful in evaluating a small number of project alternatives at a fixed location (e.g., alternative alignments for a single corridor), it has major limitations when applied to a large-scale (e.g., statewide) highway construction program. This paper discusses some of the philosophical and practical limitations of the cost-benefit analysis and how the need-based approach overcomes these limitations. A successful implementation of the need-based approach to Kansas's statewide highway improvement program is also described.
Article
Pavement management systems have been implemented across the world by transportation agencies in recent decades. To support these applications, increasingly sophisticated methods have been developed to model pavement deterioration and solve for optimal management strategies. Recently a simpler optimization approach for the system-level resurfacing problem has been developed, which is bottom-up rather than top-down, preserving facility-specific features to develop informative budget allocation results (Sathaye and Madanat, 2011). In this paper we expand upon and enhance this approach for application to large-scale, heterogeneous road networks. The methodological enhancements allow for the approach to be applied to a broader spectrum of real-world problems and efficiently for large-scale networks. These developments are implemented and presented in a case study which uses empirical models. The results are shown to be robust to deterioration model uncertainty, which is consistent with previous findings for small networks and single facilities.
Article
This article discusses how to efficiently and completely solve a bi-objective pavement maintenance and rehabilitation-scheduling problem, which aims at optimizing two objectives of pavement condition improvement and budget utilization in a simultaneous manner. This problem may be addressed by the weighting method, constraint method, ranking method, and various metaheuristic methods. However, none of these methods can guarantee the complete Pareto-optimal solution set, which would potentially lead to suboptimal decisions. In this article, a parametric method is suggested to solve the bi-objective pavement maintenance and rehabilitation-scheduling problem. The effectiveness and efficiency of the parametric method is investigated and demonstrated through a case study using the real-world data set from the Dallas District's Pavement Management Information System. A performance comparison between the widely used weighting method and the parametric method clearly justifies the computational advantages of the parametric method.
Article
The network-level infrastructure management problem involves selecting and scheduling maintenance, repair, and rehabilitation (MR&R) activities on networks of infrastructure facilities so as to maintain the level of service provided by the network in a cost-effective manner. This problem is frequently formulated as a Markov decision problem (MDP) solved via linear programming (LP). The conditions of facilities are represented by elements of discrete condition rating sets, and transition probabilities are employed to describe deterioration processes. Epistemic and parametric uncertainties not considered within the standard MDP/LP framework are associated with the transition probabilities used in infrastructure management optimization routines. This paper contrasts the expected costs incurred when model uncertainty is ignored with those incurred when this uncertainty is explicitly considered using robust optimization. A case study involving a network-level pavement management MDP/LP problem demonstrates how explicitly considering uncertainty may limit worst-case MR&R expenditures. The methods and results can also be used to identify the costs of uncertainty in transition probability matrices used in infrastructure management systems.
Article
Infrastructure systems in the US are in urgent need of maintenance and rehabilitation. According to the most recent factsheet published by the American Society of Civil Engineers, one of the top five infrastructure concerns of today in the US is the transportation system. The major challenge facing maintenance managers in state Departments of Transportation (DOTs) today is to preserve the road networks at an acceptable level of serviceability subject to the stringent yearly maintenance and rehabilitation (M&R) budgets. Maintenance managers must allocate such limited budgets among competing alternatives. Absence of simpler decision-making tools exacerbates the matter. This paper presents the development and implementation of a simpler, yet useful, network-level pavement maintenance optimization model, which is a Linear Program (LP) subject to budget constraints and the agencies’ pavement performance goals in terms of total lane-miles in each pavement condition state. A decision-making tool is developed using Frontline Systems’ Risk Solver Platform add-in for Microsoft Office Excel. This decision-making tool can compute the optimal amount of investment for each pavement treatment type in a given funding period. Pavement condition data pertaining to one of the Districts within a state DOT is used to test the model presented herein. Within this context, nine treatment types along with their corresponding unit prices ($/Lane-Mile), five pavement condition states, pavement deterioration rates, network-level pavement performance targets, and available annual maintenance budget for a 15-year planning horizon are defined. The results presented show how an annual highway maintenance budget needs to be allocated or determined to achieve the District’s value proposition for various scenarios. Comparing the results of these varying scenarios provides insight on long-term strategies and the impact of target constraints on budget expenditures.
Article
With regard to limits in budget, inevitably, one must choose prioritized projects in pavements rehabilitation and maintenance process. This paper looks into prioritization based upon a model including all effects of important factors like pavement condition index, traffic volume, road width as well as rehabilitation and maintenance cost. Since defining a model that introduces all those factors was difficult, a more advanced modeling named fuzzy logic was referred for the problem of prioritization. Although analytical hierarchy process can be used for decision making process as well, fuzzy modeling lets one have more precise choices for the outcome. Finally with the help of MATLAB software and coded M-files, inference engines such as Product engine, Dienes–Rescher and Lukasiewicz were all tested and the logical favorite separation for this application was found in product inference engine. As a case study some streets located in district No. 6 of Tehran municipality were selected and the favorite mathematical model was executed on those streets. This model was used for prioritizing these 131 sections.
Planning Pavement Maintenance and 33 Rehabilitation Projects in the New Pavement Management System in Texas
  • F Hong
  • E Perrone
  • M Mikhail
  • A Eltahan
Hong, F., E. Perrone, M. Mikhail, and A. Eltahan. Planning Pavement Maintenance and 33 Rehabilitation Projects in the New Pavement Management System in Texas. 2017. 34 12.
Decision-Making Problem for Managing Pavement 41 Maintenance and Rehabilitation Projects
  • F Wang
  • Z Zhang
  • R Machemehl
Wang, F., Z. Zhang, and R. Machemehl. Decision-Making Problem for Managing Pavement 41 Maintenance and Rehabilitation Projects. Transportation Research Record: Journal of the 42 Transportation Research Board, Vol. 1853, No. 3, 2003, pp. 21-28. https://doi.org/10.3141/1853-43 03. 44 15.
Capturing Interdependencies and Heterogeneity in the Management of Multifacility Transportation Infrastructure Systems
  • P L Durango-Cohen
  • P Sarutipand
Computer-Aided Civil and Infrastructure Engineering, Vol. 27, No. 4, 2012, pp. 278-287. 49 https://doi.org/10.1111/j.1467-8667.2011.00733.x. 50 17. Durango-Cohen, P. L., and P. Sarutipand. Capturing Interdependencies and Heterogeneity in the Management of Multifacility Transportation Infrastructure Systems. Journal of Infrastructure
Mechanistic Approach to 29 Pavement-Vehicle Interaction and Its Impact on Life-Cycle Assessment
  • O Swei
  • J Gregory
  • R Kirchain
Hindawi Publishing Corporation The Scientific World Journal Article, Vol. Volume 201, No. ID 24 524329, 2014, p. 11. 25 26. Swei, O., J. Gregory, and R. Kirchain. Probabilistic Approach for Long-Run Price Projections : 26 Case Study of Concrete and Asphalt. J.Constr.Eng.Manage, 2016, pp. 1-11. 27 https://doi.org/10.1061/(ASCE)CO.1943-7862.0001211. 28 27. Akbarian, M., S. S. Moeini-Ardakani, F.-J. Ulm, and M. Nazzal. Mechanistic Approach to 29 Pavement-Vehicle Interaction and Its Impact on Life-Cycle Assessment. Transportation Research 30 Record: Journal of the Transportation Research Board, Vol. 2306, No. 1, 2012, pp. 171-179.
A New Technique for Using Multivariate 33
  • N O Attoh-Okine
  • S Mensah
  • M Nawaiseh
https://doi.org/10.3141/2306-20. 32 28. Attoh-Okine, N. O., S. Mensah, and M. Nawaiseh. A New Technique for Using Multivariate 33
Adaptive Regression Splines (MARS) in Pavement Roughness Prediction
  • C Plati
Adaptive Regression Splines (MARS) in Pavement Roughness Prediction. Proceedings of the 34 Institution of Civil Engineers -Transport, Vol. 156, No. 1, 2003, pp. 51-55. 35 29. Loizos, a., and C. Plati. An Alternative Approach to Pavement Roughness Evaluation. 36 International Journal of Pavement Engineering, Vol. 9, No. 1, 2008, pp. 69-78. 37 https://doi.org/10.1080/10298430600949894. 38 30.
Network-Level Pavement Roughness 41 Prediction Model for Rehabilitation Recommendations
  • S Terzi
  • N Kargah-Ostadi
  • S M Stoffels
  • N Tabatabaee
Terzi, S. Modeling for Pavement Roughness Using ANFIS Approach. Advances in Engineering 39 Software, Vol. 57, 2013, pp. 59-64. https://doi.org/10.1177/0019793915592375. 40 31. Kargah-Ostadi, N., S. M. Stoffels, and N. Tabatabaee. Network-Level Pavement Roughness 41 Prediction Model for Rehabilitation Recommendations. Transportation Research Record: Journal 42 of the Transportation Research Board, Vol. 2155, 2010, pp. 124-133.