Click for next page ( 48

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 47
47 provided a high-level check of the deliverable tables to ver- Quality Control ify data completeness and data reasonableness, as well as a direct pavement distress comparison between automated/ The quality control plan was developed by the data collection semi-automated ratings and manual ratings from experienced service provider and includes quality control checks at all pavement raters. stages of the data collection, processing, reduction, and deliv- ery processes. Some of the quality control steps included con- The verification and validation also helped identify ran- trol and verification site testing, inter-rater consistency test- dom and systematic errors. Because several systematic errors ing, and numerous checks of data quality and completeness. (e.g., erroneous classification of a particular distress type) were identified, the service provider had to adjust the process Quality Acceptance and reanalyze specific types of sections. For example, the process identified problems with the rating of patches in the ODOT initially instituted additional quality acceptance checks, jointed reinforced concrete pavement sections and in the clas- which are applied to the data submitted by the contactor and sification of cracks in asphalt pavements. These problems include the following: required adjustments in the data analysis criteria. The effect of these adjustments had a very significant effect on the clas- Control site testing to help identify factors that could sification of the pavements in the various condition cate- affect the accuracy and repeatability of sensor data mea- gories, the number of deficient pavements, and the subse- surements and evaluate the quality of the collected video. quent estimated budget needs, as was discussed previously in Checks of distress ratings on batches of submitted the synthesis. data using a modified version of the service provider's distress rating software. Because these distress rating In the latest completely audited available survey (corre- checks proved to be very time-consuming and labor- sponding to 2006), using a sample of 5% of each deliverable, intensive, ODOT contracted the review of the distress the independent verification found that the percentage of ratings for the third year of collection to a consultant. the distress data meeting the tolerance requirements varied Additional data quality assurance checks of every between 93% and 98% for the various deliverables. The data element in the pavement condition database. independent quality auditor also compared the repeatability of each vehicle used by the service provider, and repro- After 3 years of consistently instituting more checks, the ducibility between the two service provider's devices and agency developed an automated procedure to rapidly and VDOT's profiler and performed a high-level data review for efficiently check the data delivered by the service provider. reasonableness and completeness (98). Figure 22 presents the main screen for the Visual Basic qual- ity acceptance tool developed within the Access database. OKLAHOMA The software tool automates the following four groups of As the Oklahoma (ODOT) started implementing a PMS, qual- checks: ity pavement condition data were identified as a key compo- nent. The agency recognized the importance of checking the Preliminary checks verify a variety of essential "gen- quality of data before they are used for important manage- eral" information included in the condition database. ment decisions and has implemented detailed quality control This step checks the district number, type of data entered and acceptance processes. in each field (e.g., integer versus characters), general section identification data, GPS values, pavement type, events (bridges, etc.), geometric values, and missing Pavement Data Collection data, among others. Sensor checks for all sensor-related data elements (i.e., ODOT established a 4-year contract with a data collection those data elements collected using lasers or sensors to service provider to collect network-level sensor, geometric, determine properties of the pavement section) look for and distress data by automated data collection techniques. duplicate records in adjacent sections, date, number of The data are processed using a combination of automated sensors used for rutting, and out-of-range values for IRI, and semi-automated techniques. Data on roughly half of the rutting, faulting, and macrotexture. network are collected each year of the contract. The con- Distress checks verify the specific distress for a given tract includes sensor data (IRI, rutting, faulting, and macro- surface type to confirm that they are in accordance with texture), distress ratings (type and severity) based on visual ODOT distress rating protocols and within the expected analysis of pavement video, and geometric data (longitudinal values not only on an individual basis but also when slope, crossfall, horizontal curve radii, and GPS coordinates). considering various distresses in combination with one Data are collected over the entire length of each section (i.e., another. sampling is not used) and reported in 0.01-mile (161-m) Special checks include more specific elements such as increments (52). maximum asphalt concrete patch length, number of