|FHWA > Policy Information > Travel Monitoring > Quality Control Procedures... > Introduction|
Quality Control Procedures for Archived Operations Traffic Data: Synthesis of Practice and Recommendations
Many traffic operations centers collect traffic data that is also archived for off-line analytical purposes, such as transportation planning, congestion monitoring, and performance measures. A survey of traffic data programs in all 50 State departments of transportation (DOTs) was conducted in 2005 to assess current practices in using archived operations data for planning purposes.1 This survey found that unknown or inadequate data quality was one of the most frequently mentioned barriers to integrating archived operations data into traditional statewide traffic databases. Other published reports and workshops have identified similar concerns with the quality of archived operations data.2,3
The quality of archived data from traffic operations systems has been influenced by two prevailing issues:
The result has been that many data archive managers and users have wrestled with data quality problems.
Traffic operations managers use data for different applications than transportation planners, and as such, may have different data quality requirements. Because of these different data uses, many transportation planners are unsure or skeptical of operations data quality, sometimes without the benefit of any actual data comparisons. In other cases, data comparisons and accuracy evaluations have been conducted and have raised concerns about the quality of archived data for historical/analytical applications.
In summary, the quality of traffic data archived from some traffic operations centers has been identified as a concern. Using quality control procedures to monitor and identify traffic data quality problems is an essential component of improving data quality. A 2003 FHWA report, Traffic Data Quality Workshop: Proceedings and Action Plan,2 recommended that quality control procedures be compiled and synthesized in a guidance document. This report is intended to meet this recommendation for a guidance document.
Quality Control vs. Quality Assurance
This report addresses quality control procedures for archived traffic data. In this report, quality control is defined as the identification, review, and treatment of data that does not meet specified acceptance criteria. Quality control and quality assurance are often used interchangeably to mean the same thing; however, the terms as used in this report have different meanings with important implications.
The term "quality assurance" is used to encompass actions taken throughout the entire traffic monitoring cycle to ensure that traffic data meet or exceed customer expectations. With this definition, quality assurance includes actions taken before data collection as well as after data summarization, such as the following:
The term "quality control" is used to describe the process performed after data collection but prior to data summarization and reporting to ensure that traffic data meet certain acceptance criteria. As used in this report, quality control involves reviewing and manipulating data that have already been collected. Data quality actions that are restricted to simply fixing data that have already been collected (referred to as "scrap-and-rework") are ineffective in the long-term because they address the symptom but not the root cause of poor data quality.
However, in some cases, users of archived data may have little or no direct control over quality assurance processes other than quality control. For example, planning groups that wish to use archived data may currently have little to no input on procuring and testing operations-based traffic sensors. At the same time, though, users of archived data should recognize that long-term improvements to poor data quality can only be made by moving toward a position that influences how real-time traffic data are collected and archived.
Overview of this Report
With these long-term quality assurance goals in mind, though, this report will focus only on quality control procedures, or the review of archived data that have already been collected and saved by traffic operations systems. This report summarizes quality control procedures used in numerous archived data management systems (ADMS) implementations. This report provides recommendations for a basic set of quality control procedures that can be adopted, as well as a process to customize quality control procedures for system-specific data quality issues. This report also details the typical steps involved in quality control procedures, including the automation of quality checks, the use of manual visual review, the flagging of failed data records, and the use of metadata to document quality control actions.
Microscopic vs. Macroscopic Validity Criteria
Because of its focus on archived operations data, this report focuses on "microscopic" validity criteria for traffic data summarized for relatively short time intervals (typically 5 minutes or less). Microscopic validity criteria are intended for use on original source data that have been obtained directly from a traffic management system. These microscopic validity criteria are contrasted with macroscopic criteria, which are more commonly associated with the longer time intervals (typically hourly intervals or greater) used in planning-based traffic databases. An example of macroscopic criteria is the following: The volume total in any lane for the entire day must not be 0. In many cases, archived data must be summarized to longer time intervals before macroscopic criteria can be applied. Additionally, a summary of macroscopic validity criteria can be found in other resources.4,5
The intended audience for this report includes developers of ADMS as well as researchers and analysts of archived data. Software developers can adapt some or all of the quality control procedures as deemed necessary. The report could also be used or specified in writing a request for proposals to develop an ADMS. Researchers and data analysts can use the report to perform additional quality control if data quality remains an issue.
1 NCHRP Project 7-16 Interim Report Recommended Revisions to the AASHTO Guidelines for Traffic Data Programs, National Cooperative Highway Research Program (NCHRP), April 2006.