Pavement management systems in local governments

This study investigates the use of real‐world, manually collected historical pavement condition survey data from local government pavement management systems (PMS) to develop robust performance models and life cycle cost analyses (LCCA). The research aims to harness existing datasets from systems such as StreetSaver and MicroPAVER and convert raw distress information into actionable insights for treatment selection, maintenance scheduling, and long‐term budgeting. Central to this doctoral dissertation is the effort to generate unified, probabilistic performance models that forecast pavement deterioration. In doing so, the study not only highlights the potential for improved localized understanding of treatment performance but also bridges the gap between empirical condition data and economic decision-making frameworks.Literature showed that PMSs rely overly on the condition indices and not the data gathered to calculate the index. Relying solely on the Pavement Condition Index (PCI) for decision-making neglects the rich, detailed distress data originally used to calculate it. This approach discards valuable information that could otherwise enhance decision support. Moreover, many local agencies lack comprehensive historical records of construction activities due to inadequate as-built documentation. Additionally, pavement management systems frequently overlook age-related cracking—a common issue in all asphalt pavements. By utilizing proactive performance models to time preventive treatments, agencies can slow deterioration more cost-effectively than by adopting reactive strategies that only address damage after cracking becomes visible.Equally concerning is the insufficient focus on fatigue cracking, which compromises pavement structure and typically occurs in areas with heavy traffic. Without forward-looking estimates to predict the onset of fatigue cracking, management strategies become reactive, postponing essential interventions until distress is already apparent. This is particularly problematic given that reflective cracking, which can progress rapidly under temperature fluctuations and heavy loads, may also develop. Furthermore, local governments often depend on default performance models that define pavement “life” solely as the time until failure, without updating or validating these models with their own data. Treatment selection within decision trees is similarly unexamined, leading to inadequate differentiation between strategies for age-related cracking— where proactive crack sealing is crucial—and fatigue cracking, which demands structural reinforcement.The methodology begins with a comprehensive review and extraction of detailed distress data, including distress type, severity, and extent, from historical condition surveys. Recognizing that the core indicators of pavement performance are embedded within these details rather than in the aggregated Pavement Condition Index (PCI), that is commonly used as the only engineering metric for local streets and roads, the study employs a transformation process. This process uses a uniform response variable—Deduct Value (DV)—to correlate and convert disparate measures into a single, interoperable metric. This conversion is critical as it allows the study to overcome the inherent mathematical incompatibilities in raw data, thereby enabling the creation of probabilistic survival curves. These curves estimate the time intervals required for pavements to reach various thresholds of cracking, an essential step in predicting the onset of significant pavement deterioration.Beyond data transformation, the study’s performance modeling incorporates variables that are often overlooked in standard PMS databases. It examines how factors such as functional class differences, traffic volume (including the presence of bus and truck routes), climatic conditions, and local construction practices influence treatment performance. For example, the analysis reveals that the existence of bus routes can reduce the effective life of pavement treatments by approxima ely 12–15%, and in some cases, even accelerate age-related cracking, most likely reflection of age-related cracks in existing underlying layers, by 20–25%. Such insights underscore the necessity of incorporating agency-specific variables into performance models to more accurately forecast future maintenance needs.A significant contribution of this research lies in its integration of performance modeling with LCCA to provide a comprehensive, long-term economic perspective on pavement maintenance and rehabilitation (M&R) strategies. The LCCA framework developed here is built on two pivotal components: the cost of treatment and the timing of treatment interventions. The study innovatively employs clustering techniques to account for economies of scale in treatment costs, drawing upon historical cost data—specifically, an inflation-adjusted Caltrans cost database—to determine unit cost variations across different project volumes. By segmenting data into clusters using k-Means and Hierarchical clustering methodologies, the research delineates clear thresholds that capture the cost–volume relationship. Within each cluster, Monte-Carlo simulations are used to address the variability of cost data. This simulation-based approach generates a range of potential cost outcomes rather than a single deterministic figure, thereby accommodating variations due to contractor differences, geographic considerations, seasonal factors, and other project-specific deviations.The integration of probabilistic performance models with LCCA allows the study to explore the economic implications of various M&R strategies. For instance, one comparative case study contrasts preventative maintenance with continuous rehabilitation strategies. Although the cost difference between the two approaches may be marginal—often less than 10% in deterministic terms—the model emphasizes that the additional road-user costs, safety implications, and accelerated deterioration associated with continuous rehabilitation render preventative maintenance a more desirable policy alternative in the long term. Another case study focuses on contrasting fatigue-induced cracking with age-related distresses, demonstrating that fatigue cracking, due to its progressive nature and the resultant full-depth pavement failure, entails significantly higher life cycle costs.The research further explores policy impacts through LCCA, illustrating how modifications to operational practices can yield substantial savings over the pavement’s life cycle. One notable example is the analysis of asphalt compaction requirements. By enforcing a higher mandated compaction threshold, the study shows that local agencies could realize up to 20% in life cycle savings. Such findings not only offer empirical support for potential policy changes but also serve as a compelling argument for the broader adoption of performance-based maintenance strategies.Additionally, the study sheds light on several challenges currently facing local agencies in PMS data management and treatment tracking. Inconsistencies in maintaining historical survey data, especially due to changes in street and section identification formats, have led to significant data discontinuities. Manual condition surveys, which typically inspect only a representative sample of each pavement section, further complicate the accurate tracking of distress progression and treatment impacts. Lack of accurate as-builts in agency PMSs is another challenge as inaccurate or incomplete M&R history data can weaken the dataset for various distresses. The absence of a robust mechanism to account for “digouts” or “base repairs” exacerbates the difficulty in accurately correlating treatment performance with subsequent distress patterns.In summary, this study provides a detailed and integrated framework for leveraging historical PMS data to predict pavement performance and optimize maintenance strategies through probabilistic modeling and life cycle cost analysis. By combining engineering performance models with economic analysis, the research ffers an approach that not only improves the scientific understanding of pavement distress mechanisms but also informs policy decisions that can lead to substantial cost savings and enhanced pavement longevity.