<< Previous

Contents

Next >>

Advanced Quality Systems: Guidelines for Establishing and Maintaining Construction Quality Databases

CHAPTER 2. LITERATURE REVIEW

2.1 INTRODUCTION

As the first step in the research, a comprehensive literature search and review was performed to obtain information on construction quality database systems used nationwide. The search included national (Transportation Research Information Service [TRIS], Research in Progress [RIP], FHWA, TRB) and state agency databases. Many published documents on construction/materials quality measures, processes, specifications, and systems were identified and retrieved for in-depth review. Appendix A contains a bibliography of all the literature collected and examined in the study.

This chapter presents a summary of the information deemed most pertinent to this study. While a considerable amount of good information has been written in recent years regarding the development and use of QA practices and programs (i.e., QC, acceptance, IA, PRS), the focus of this search/review was on quality database systems. Of particular interest was information on data items, database schema and architecture, data entry and security, data querying and analysis tools, and integration with other databases.

2.2 LITERATURE REVIEW FINDINGS

Construction quality databases can be defined as computerized databases containing a variety of pavement construction-related data that characterize the quality of materials and workmanship used to construct the pavement. The primary purpose of a construction quality database is to facilitate the assessment of quality of materials production and placement, including the establishment of pay factors, as defined by specifications. A secondary purpose is to enable detailed research analyses of quality, performance, and cost data that can help guide future improvements to current standards and specifications (e.g., materials, techniques, and design strategies to use; quality characteristics and levels to use in acceptance; incentive/disincentive plan).

While most SHAs have established and maintained construction-related databases for many years, it is only within the last 5 to 10 years that they have realized the need for more detailed systems to accommodate the requirements of new QA programs (developed in response to Title 23, Part 637, Code of Federal Regulations [FHWA, 1995]). The exact number of construction quality databases currently in existence at the State level could not be ascertained through the literature review. However, based on the results of a recent National Cooperative Highway Research Program (NCHRP) survey on State construction QA programs (Hughes, 2005), it is believed that several States have some form of a quality database in use, particularly as it relates to hot-mix asphalt (HMA) and portland cement concrete (PCC) paving materials. This survey, which among other things focused on QA programs for soils and embankments, aggregate base and subbase, HMA paving, PCC paving, and PCC structures, indicated that all of the 45 responding agencies (43 States, the District of Columbia, and the FHWA Federal Lands Division) have a QA program in place for HMA paving, while about two-thirds have such a program for PCC paving.

Presented below are discussions of works undertaken in recent years having special focus on the various attributes of construction quality databases. Information gleaned from the following documents was instrumental in the development of the concepts and best practices reported herein:

• NCHRP Synthesis 346, "State Construction Quality Assurance Programs" (Hughes, 2005).
• FHWA Manual, "Optimal Procedures for Quality Assurance Specifications" (Burati et al., 2003).
• FHWA Report Evaluation of Procedures for Quality Assurance Specifications (Burati et al., 2004).
• Transportation Research Circular Number E-C074, "Glossary of Highway Quality Assurance Terms" (TRB, 2005).

2.2.1 Construction Quality Data Items

Results of the aforementioned NCHRP survey (Hughes, 2005) showed a variety of attributes being used for QC and acceptance of HMA and PCC paving materials. Table 1 shows gradation, asphalt content, volumetric properties, and compaction as the most frequently used QC attributes for HMA. These same characteristics and ride quality are also most common for acceptance of HMA. For PCC paving, table 2 shows gradation, air content, and slump as the most common QC attributes, and thickness, air content, cylinder strength, slump, and gradation as the most common acceptance characteristics.

Table 1. Attributes used for QC and acceptance of HMA paving (Hughes, 2005).

Attribute	Number of responses
	QC	Acceptance
Asphalt content	40	40
Gradation	43	33
Compaction	28	44
Ride quality	16	39
Voids total mix	20	26
Voids in mineral aggregate	26	23
Aggregate fractured faces	25	23
Thickness	13	22
Voids filled with asphalt	19	13

Note:

44 total responses

Table 2. Attributes used for QC and acceptance of PCC paving (Hughes, 2005).

Attribute	Number of responses
	QC	Acceptance
Air content	25	38
Thickness	14	36
Slump	24	33
Cylinder strength	18	31
Gradation	25	26
Beam strength	14	18
Water-cement ratio	12	16
Ride quality	1	15
Aggregate fractured faces	7	6
Sand equivalence	0	3
Permeability	0	3
Core strength	0	2

Note:

44 total responses

A review of State specifications by Burati et al. (2004) indicated similar findings with regard to quality characteristics used in the QA program. This review also showed gradation, asphalt content, air voids, in-place density, and smoothness as the quality characteristics used in determining pay factors for HMA, and thickness, air content, smoothness, and flexural or compressive strength as the quality characteristics for determining PCC pay factors.

2.2.2 Web-Based Quality Systems

The following is a review of two Web-based quality systems, HMA View and ELVIS.

HMA View

The development of a Web-based HMA database system, HMA View, by the University of Washington, and its use was the focus of a recent paper by White et al. (2002). The motivation for developing the system was the desire to (a) link all phases of a project life–cycle—mix design, construction, usage/maintenance, and performance—to allow for the evaluation of Superpave mixes and (b) bring current computing technology and HMA construction together to create real-time (or near real-time) construction tools.

A schematic of the Web-based, server/client-type database system is provided in figure 1. The system can open different avenues of data entry to each of the users and allows a variety of users from various locations access to a large-scale data warehouse in real time. It supports a wide variety of data types, including test results, digital and infrared images, inspection videos, instrument readings, and audio clips, and each type of media can be referenced to a specific location using global positioning system (GPS) technology.

Figure 1. Schematic of HMA View system (White et al., 2002).

Brief descriptions of the system’s features and capabilities are given below.

• Data Acquisition—Data may be entered directly into the database through data entry points on the Internet or imported electronically from other data sources. Manual entry is facilitated with secure Web forms customized for each user group. Electronic entry is done through formatting and the use of simple translators. As data are entered into the database, they are validated and checked against rules to ensure that any later analyses will be performed on a consistent data set.
• Data Warehousing—To provide users with the ability to browse and analyze the collected data, an efficient storage method was devised whereby data are stored and cataloged into a specific location for easy retrieval. The data warehouse includes data mining capabilities.
• Browsing—To serve the needs of multiple users interested in differing "views" of the data, the following five browsing modes are included:
  • Project Info—Quick snapshots of the project, including spatial view.
  • Mix Design—Detail displays regarding the job mix formulas (JMFs) used in the project.
  • Construction—Multiple subsections including general site data, HMA data (gradations, volumetrics), density data, and media content.
  • Performance—Historical performance data, as well as traffic and usage statistics.
  • Analysis—Allows creation of plots and performing diagnostics on different construction parameters.
• Queries—Advanced queries can be performed via parameterized Boolean methods (e.g., projects paved in 1999 with an ambient air temperature of less than 59°F) or spatial methods (i.e., visual search for a project based on location on a map).
• Analysis—Both intra- and inter-project analysis capabilities are available by means of graphing (QC charts) and table generation.
• Data Export—Data may be exported into spreadsheet format for more in-depth analyses than what the database allows (e.g., statistical analyses).

The HMA View system described by White et al. (2002) includes data backup capabilities and maintains transaction logs. It also includes secured (login and password scheme) client portals for entering and viewing data.

ELVIS

Yuan et al. (2006) reported on the implementation of a Web-based electronic data management system (EDMS) for material QA on a large highway construction project in Texas. The on-going 49-mile design-build toll road project, located on SH 130 southeast of Austin, utilizes an independent construction QA firm for administering QA functions and making acceptance decisions. The QA firm uses a specially designed Electronic Document Management System (EDMS), called Electronic Laboratory Verification Information System (ELVIS), first implemented in 2004.

ELVIS was developed on a Microsoft^® Net platform, is rooted in Structural Query Language (SQL) database, and runs on a Microsoft^® Windows 2003 web server shielded by a firewall. As illustrated in figure 2, it consists of eight key functional components. Brief descriptions of each are provided below:

• Integrated database management system (DBMS)—The DBMS is a set of inter-related database tables based on technology that permits the QA firm staff to organize, store, and manage data related to all planned construction quality assurance activities. Data stored in the DBMS include sample identification information, test procedure-based testing data, material specification requirements, PCC and HMA proportion designs, nonconformance records and disposition information, controlled vocabulary languages (CVLs), and encrypted user access data.
  • CVLs are sets of pre-determined terms that are used consistently to describe certain materials, work features, and technical requirements. CVLs are essential in searching, locating, retrieving, and grouping information by linking similar records and resources with a combination of unique terms. CVLs used in the QA program include technician/inspector names, sample type, report type, material/mix code, supplier/producer, specification item, special provision, material grade/class/type, roadway, direction, feature, structure number, and sieve size.
    

    Figure 2. Schematic of ELVIS system architecture (Yuan et al., 2006).

    
  • PCC mixture designs stored in the database include batch plant information, mix design code, concrete class, design and specification requirements, batch weights per cubic yard, admixtures, and approval information. A set of integrated database tables is dedicated to store HMA mixture design data, including batch plant information, mix code, grade, asphalt binder information, sources of fine and coarse aggregates, aggregate proportion percentages, JMFs, and approval information. JMF records for each mix design are chronologically indexed in the database for future retrieval.
• Workflow-driven data management functions—Workflow technology was employed to facilitate data management and information process automation. The data management workflows in ELVIS consist of three integrated components: data entry workflows, engineer review/authorization workflows, and document control workflows. The system uses Crystal Reports^® software to publish database-driven reports to the Web. These reports can be exported to electronic formats (e.g., .pdf files) used by most end users. ELVIS supports 43 field and laboratory test procedures for a wide range of highway construction materials, including soil and base materials, PCC, HMA, and coarse and fine aggregates for PCC and HMA mixtures. The system is capable of determining pay adjustments for HMA pavements based on deviations of laboratory molded densities and the percent of in-place air voids, and for pavement ride quality based on the measured International Roughness Index (IRI).
• Material baseline information management applications—ELVIS workflows automatically archive proctor curve information, including auto-generated numerical Proctor curve identification number, curve equation parameters, maximum dry density, optimum moisture content, Atterberg limits, percent minus the No. 40 and No. 200 sieves, material description based on laboratory classification, material source, and sample location in the database when Proctor curve tests are authorized.
• Online information delivery system—As a Web-based EDMS, ELVIS serves as a secure, real-time, common-information sharing platform among a broad constituency of users, including managers, engineers, QC/QA technicians and inspectors, construction superintendents, material vendors, and designers. In addition to standardized Crystal Reports^®, which can be downloaded, printed, and emailed as a .pdf or image file, the system allows users to generate summary data reports in Microsoft Excel^® format.
• Engineering decision tracking functions—Depending on the severity of material nonconformance, the independent QA engineers may accept certain out-of-specification material tests based on engineering discretion and the Engineer Decision delegation agreed upon by the Texas Department of Transportation (TxDOT). ELVIS automatically tracks and documents out-of-spec materials and test reports, and engineering justifications for acceptance decisions.
• Deficiency management—ELVIS is also capable of tracking and monitoring failing material tests, and retests on the reworked products and materials. Using database techniques, the system associates retests with the original failing test. Key information and status of original failing tests and their retests are posted automatically on a Web page. Based on real-time construction deficiency information, QC staff and field engineers can plan and monitor corrective measures. When a construction deficiency has been corrected or reworked and the product has passed retests, QA engineers can close the deficiency by changing the deficiency status from "Pending" to "Closed" through a secure Web page.
• Statistical analysis tools—The application of statistical analysis to the test history of construction materials is used as a guide to identify the material variability and to develop processes to control the variability of manufactured materials (e.g., PCC, HMA, flexible base, and aggregates/products for PCC and HMA mixtures). A series of Web-based statistical analysis tools are built into ELVIS, with the system being capable of determining such statistics as mean, variance, range, standard deviation, deviation significance (i.e., number of standard deviation), and running averages.
• System administration tools—An IA program is part of the SH 130 project. Administrated by TxDOT, the IA program evaluates all sampling and testing procedures, personnel, and equipment used by the independent QA firm in making acceptance decisions. To ensure personnel qualifications, ELVIS documents and tracks technician certification information (e.g., who is certified by the IA program to perform specific test procedures and the expiration dates of technician certifications on various test methods). Similarly, equipment calibration records and calibration schedules are stored in the system.

Yuan et al. (2006) noted that successful implementation of ELVIS has provided a cooperative material test data processing platform and enabled the processing and reporting of up to 400 material acceptance tests per day. All QA test reports are stored systematically in the ELVIS database, and authorized users can use secured online access to retrieve and produce any test report in the flexible data format in seconds. Reportedly, the rapid reporting of test results has greatly helped in the management of material quality-related construction deficiencies and improved the builder’s quality performance.

2.2.3 Integration of Pavement Management and Other Systems With Construction Quality Databases

In a 2003 study, Hudson et al. examined the question of how existing pavement management data and construction/materials data can be used to evaluate the performance of new materials and techniques, and to validate new design methods. In the first of two phases, visits were made to five States (Maryland, Indiana, Florida, Arizona, and Washington) to discuss aspects of their PMS’s and Superpave materials data, and to examine linkages between materials/construction data and pavement management data. Key findings included:

• Absence of a convenient link between essential data on materials characteristics and PMS performance data. Most often caused by the fact that materials/construction data are (a) commonly stored in flat files, (b) difficult to access, and (c) sometimes incomplete.
• Valid analysis of the performance of any design, material, or technique can only be done when relevant data are available electronically.
• Performance data can only be linked to materials/construction data when use is made of a common locator reference.

Hudson et al. (2003) noted that failure is caused by many factors (e.g., mix composition, mix temperature at time of construction, degree of compaction, actual thickness, subgrade properties, drainage problems, high traffic loads) applicable to the lot where failure occurred. Data are not always complete (e.g., actual thickness is often not recorded) or they are difficult to retrieve. As a result, the failed lot cannot be traced to its materials/construction properties. The current bestcase scenario is that performance data are averaged over a mile and are compared with average material/construction property values over that mile. In many cases, the data will only allow such a comparison for the average properties for the entire project. Few records are kept about the exact location of lots or sublots. Thus, it is imperative to have the ability to compare performance and materials/construction characteristics on a lot-by-lot basis.

Easy access to all information required for judging the performance of pavements and materials can only be realized when the following conditions are met:

• All data required to evaluate performance of materials, techniques, etc. have to be available in electronic format.
• All individual performance and materials/construction data should be tied to their exact locations, either in divisions of milepoints or in geographic information system (GIS) coordinates.
• The files with these data should be made accessible to the users in the organization that need such data to work. A convenient way to do this is to run these files on servers with proper protection measures. This concept can be extended with the possibility of loading the required data into a web-based system that can work as a data warehouse, data viewer, or data sorter.

Hudson et al. (2003) recommended that relevant data from the materials/construction database and PMS database be made available and transferred electronically to a performance analysis database. This third database can be made of a commercially available spreadsheet system (e.g., Excel^® or Lotus) or a Web-based system that extracts relevant information and makes overviews, graphs, and reports.

Key to linking databases, as shown in figure 3, is to have precise, unambiguous location identification and date/time information. Unambiguous locations can be provided by GPS measurements, but they must be tied to traditional location identification information, such as project number, milepoint, lane, direction, date, etc.

Figure 3. The concept of linking databases (Hudson et al., 2003).

The concept of integrating materials/construction and PMS data was tested in Phase II (nicknamed the Pathfinder study) using seven Superpave projects in Maryland. As a first step in this evaluation, data fields from the QA, pavement design, mix design, and PMS files were identified. The second step involved collecting all the required data for the seven Superpave mixes and entering them into separate electronic databases. The third step involved compiling data into a suitable database for storage, linking, analysis, and reporting, using the HMA View Web-based system. The final step involved performing various types of performance analyses.

Results of the Pathfinder study showed that, despite being a cumbersome and time-consuming process, it is possible for a SHA to assemble a database that can be used to evaluate performance of Superpave and other designs, materials, and techniques. It was also learned that there are much more data present in pavement management, pavement design, materials and construction files than are currently used (or accessible) for performance analysis, and that some of the missing information could be collected easily in the future.

A recent study sponsored by the Arizona Department of Transportation (ADOT) resulted in the development of an enhanced PMS designed to help ADOT planners and engineers better develop and execute the 5-year highway construction program (Li et al., 2006). The new system is more comprehensive and more substantially integrated with other ADOT systems, including the construction materials, maintenance activities, and features inventory databases, and the Department’s deflection-based overlay thickness design procedure (using data collected with a Falling Weight Deflectometer [FWD]).

The new system was implemented using a Microsoft SQLServer^® 2000 database. The database is relational (it uses a commercial relational database management system [RDBMS]), and tables are designed based on functional dependency. The RDBMS allows various types of detailed data, such as inventory, traffic, pavement structure history, maintenance history, and pavement condition data to be recorded on their own segmentation basis. Through relational database design, all the data from various sources are consolidated and logically related by using a common highway ID and location referencing system.

Integration of the FWD-deflection based overlay thickness design procedure makes pavement designers routine users of the new system. Pavement performance is predicted using site-specific modeling with default performance class-based models (used when insufficient data are available for site-specific modeling). A functional module included in the system provides information feedback for evaluating the effectiveness of specific activities in terms of performance and cost for a specific group of sections.

Zhang and Zhou (2002) reported on the implementation of a database and information system for forensic investigation of pavements in Texas. While forensic investigation is a science focusing on the determination of the causes of premature failures, much of the work conducted in this project (and previous related projects [Victorine et al., 1997; Zhang et al., 1999]) centered around the evaluation of existing TxDOT databases and the development of a forensic system (ForenSys) capable of accessing important data contained in the other systems. The relevance of this work to construction quality databases is significant in that there’s a shared goal of utilizing construction materials and workmanship quality to help diagnose the performance of the as–constructed pavement.

The framework for the ForenSys database (figure 4) was first developed in 1998. Completion of a stable and fully implementable system followed in the years thereafter. The newly enhanced and integrated system utilizes the location reference system used in the Department’s pavement management information system (PMIS) database (see table 3) and acquires its data from the Layer database (pavement type, layer types, thicknesses, material properties, construction year) and the PMIS database (distress, ride, friction, traffic, highway geometry).

Figure 4. Conceptual framework design for TxDOT ForenSys database system (Zhang and Zhou, 2002).

Table 3. Sample location reference data in TxDOT’s ForenSys database (Zhang and Zhou, 2002).

County Number	Highway,Roadbed ID	Beginning Reference Marker	Beginning Displacement	End Reference Marker	End Displacement
127	BU0067K	0450	0.5	0450	1

The ForenSys database software makes the pavement-related data easily accessible to the forensics engineer and provides an interface to easily store, display, and analyze forensic investigation results. Additional system improvements recommended in the report include:

• Supplement relational technology with object-oriented technology to allow the system to handle photographic and other object files, in addition to mere data.
• Upgrade the software from use as a stand-alone system on a single computer to a webbased system easily accessible by authorized forensics personnel.

2.3 SUMMARY

The development and implementation of State construction QA programs in recent years, coupled with the advancements in computer technologies and PRS, has created an environment ripe with opportunities for fully integrated and Web-enabled construction quality databases. This chapter presented many of the quality concepts that have recently been studied and/or put into practice, thereby providing a vision of the future of construction quality databases. This vision is reinforced by the planned and in-progress efforts of others to establish or update current systems to an open, Web-based quality system. The features and capabilities of such a system can be summarized as follows (Benson, 2004; MTO, 2004; ConnDOT, 2006; Mrawira et al., 2002; Harvey, 2000):

• Access by a wide range of agency and non-agency users (e.g., contractors, producers, testing lab personnel, construction engineers and inspectors, researchers, planners) stationed in a variety of locations.
• Real-time use to facilitate timely processing and validation of sample data and test results.
• Various levels of information detail (storage, viewing, and reporting).
• Multiple levels of access security (roles and privileges).
• Secure transactions between users and the quality system.
• Logging and tracking of material samples, related tests, and results.
• Audit information including data entered, who entered the data, who modified the data, and reasons for data revisions.
• Electronic approvals and event notifications.
• Increased data integrity through data capture at the source (including wireless devices).
• Ability to perform rudimentary data analyses and graphing for QC, QA, and IA functions.
• Use of geo-referencing and barcoding technologies.
• Ability to interface/link with other agency systems (e.g., pavement management database, financial/cost accounting system) via fixed location identification indices, time indices, and accounting indices.
• Ability to create ad hoc and pre-defined reports in a variety of formats.
• Functionality for report distribution via e-mail or an approved website.
• Capability for system backups and disaster recovery.

<< Previous

Contents

Next >>

Events

• 9th International Conference on Concrete Pavements
  San Francisco Marriott Hotel, San Francisco, CA
  August 17-21, 2008

• View all Upcoming Pavements Events

Contact

Peter Kopac
Turner Fairbank
202-493-3151
E-mail Peter

 

This page last modified on 05/01/07