Sampling Procedure for Performance-Based Road Maintenance Evaluations

Abstract

system (1,2). This initiative calls for performance-based work, in which a desired outcome is specified rather than materials or techniques to be used by the contractors to maintain the asset items. In performance-based contracts, road administrators define performance measures that specify the minimum acceptable condition at which the asset items should be maintained. To ensure that contractors maintain the asset items according to these measures, road administrators must design and implement a comprehensive and reliable performance monitoring process. One of the most important areas in the performance monitoring process is inspections conducted in the field. Field inspections require a significant amount of planning and careful monitoring. Defining a procedure that guarantees the success of field inspections is a challenge. When defining such procedure, road administrators must consider budget and time limitations, among others. These constraints influence the final techniques adopted that address issues such as the portion of the population to be inspected, frequency at which inspections should be conducted, and methodology to be used to collect information. Since performancebased road maintenance contracts are relatively new, the availability of guidelines that address such issues (with focus on performancebased contracts) is limited (2). This paper presents a three-stage and seven-step statistical sampling procedure developed to ensure that findings from field inspections will be reliable and representative with high confidence of the actual condition of asset items in the entire population. Also, this paper presents three alternative approaches of sampling for the cases when sampling needs to be performed not just once, but multiple times over the duration of a performance-based road maintenance contract (e.g., once in every year to monitor a contractor working under the terms of a 5-year-long performancebased road maintenance contract). Of note is that this paper does not discuss the operational issues related to the condition assessment process, such as frequency of inspections and inspection techniques. Discussion of points presented is limited to the sampling aspect of the condition assessment process. RANDOM INSPECTIONSIn road maintenance evaluations, conducting a 100% inspection may not be feasible because of time and cost constraints. The condition data for a number of asset items such as pavement and shoulders can be collected by using road scanning vehicles at traveling speeds without necessarily deploying crews to perform manual assessments in the field. However, data collection with respect to many other asset items, such as pipes, ditches, underdrains, and fences, requires a considerable amount of time to be spent by the crews in the field. It is not feasible to collect the condition information for such asset items for the whole population when such information cannot be collected at traveling speeds. When collecting information for the Maintaining the road infrastructure at a high level of con...

Roadway Asset Inspection Sampling Using High‐Dimensional Clustering and Locality‐Sensitivity Hashing

“…One classic sampling method in transportation maintenance management is stratified sampling as it balances trade‐offs between inspection costs and sampling accuracy (Adams and Winkelman, ; De la Garza et al., ; Medina et al., ). De la Garza et al.…”

“…De la Garza et al. () proposed a stratified sampling method for road maintenance evaluation. The stratification criteria include geographical location, weather variation, urban and rural setting, and traffic volume.…”

A condition assessment method for roadside assets

“…Although this approach for determining sample size is relatively simple, it may not be justified statistically. In order to correctly define a sampling procedure, de la Garza, Pinero, and Ozbek (2008) suggested that the characteristics of the 'population, sample units, asset items within each sample unit, and acceptable quality levels must be understood'. Several statistical methods have been proposed in the literature for computing the number of sample units needed to be inspected (i.e.…”

“…sample size). For a given precision and confidence level, the necessary sample size is determined as a function of size of the network, estimates of the population variance, desired precision and desired confidence level (de la Garza et al, 2008;Kardian & Woodward, 1990;Medina, Haghani, & Harris, 2009). This approach for determining sample size is founded on statistical principles and is adopted in this paper.…”

Performance Based Road Asset Management System, with a case study: Abu Dhabi

“…Appropriate control procedures as well as penalties for non-compliance have to be well Researchers (De la Garza et al, 2009, Pin˜ero, 2003, Pin˜ero and De la Garza, 2004, de la Garza et al 2008de la Garza et al , 2009 identified five components in order to develop a framework for monitoring performance in PBC. These five components can be applied as guidelines to establish a performance monitoring system for PBC in Abu Dhabi.…”