Roadway Measurement System Evaluation (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

The SHRP 2 safety research plan includes two tracks: a field study of driving behavior using vol- unteer drivers and a sophisticated instrumentation package installed in the volunteers’ vehicles (in-vehicle study), and a video system to record movements of all vehicles at specific road sites such as intersections (site-based study). The in-vehicle study consists of three areas: (1) study design and field data collection, (2) roadway data, and (3) analysis. Safety Project S03, Roadway Measurement System Evaluation, fits under the roadway data area of the in-vehicle study. The goals of the Safety Project S03 research were to • Prequalify commercial data collection vendors for bidding on Safety Project S04B, Mobile Data Collection; • Evaluate the precision and accuracy of the mobile roadway and pavement inventory data col- lection services as they relate to selected roadway safety data elements collected at highway speeds; and • Produce a recommended list of roadway data elements and associated specifications to be collected under the S04B project. On December 19, 2007, SHRP 2 awarded Safety Project S03 to the research team of Applied Research Associates, Inc. (ARA), Cambridge Systematics, Inc. (CS), and KCI Technologies, Inc. (KCI). The work for this assignment was completed under the following tasks: • Finalizing the work plan and the Quality Assurance/Quality Control Plan; • Determining the prioritization of data elements; • Developing the test site selection criteria; • Formulating the test site evaluation and recommendations; • Conducting the test site mapping and surveying; • Organizing and completing the field data collection; • Evaluating the data collected; and • Producing the final report. The core research was accomplished through the development, organization, and conduct of a roadway measurement system evaluation (rodeo). Ten commercial data collection vendors and the Federal Highway Administration (FHWA) participated in the rodeo evaluation and provided the research team with preliminary data. Two types of data sets were provided to the research team: initial data set and final data set. The purpose of the initial data sets was to provide insight into the participants’ ability to collect data consistently under real-world survey conditions likely to be encountered during S04B. The purpose of the final data sets was to evaluate the participants’ Executive Summary1

2ability to collect the specific roadway assets and features of interest to the roadway safety research community. Each participant was to provide the research team with three repetitions of data col- lection for each data set. The initial data sets covered a total of 258 lane miles of rodeo survey routes and included digital images, Global Positioning System (GPS) coordinates, and roadway pro- file and geometrics data. These were delivered by the participants before leaving the rodeo site. The final data sets covered a total of 8.5 lane miles on the six specific rodeo test sites and included the roadway data elements from the data elements list. The participants postprocessed these data and delivered them after the conclusion of the rodeo. The research team evaluated the initial and final data sets provided by participants based on the data quality attributes of coverage, consistency, completeness, and accuracy. The evaluation of the initial data sets indicated that consistency of GPS data was very good for all sites, including various land use and cover conditions such as urban, rural, canopy, and highway locations. Five of the six data collection teams reporting GPS data achieved a consistency rating of sub-30 cm (1 ft). Three of the 10 commercial vendors—Teams 06, 08, and 10—were eliminated from further consideration upon evaluation of the final data sets because of their lack of coverage or data for- mat issues. Of the remaining seven teams, none provided all of the requested data elements. Some vendors focused on providing only a few data elements that were complete; others pro- vided more data elements but may have sacrificed accuracy and precision in doing so. Because the results of the rodeo were inconclusive, SHRP 2 decided to prequalify all the participants. Team 10 decided not to continue in the pursuit of Safety Project S04B. Overall, it was concluded that most of the desired data elements can be collected, although the desired target accuracy, as used in the rodeo, may not be achieved. However, it was observed that vendors were largely inconsistent in adhering to the data collection procedures and data processing requirements outlined in the rodeo data collection plan and data elements list. There appeared to be general confusion regarding engineering terminology, such as Manual on Uniform Traffic Control Devices (MUTCD) code, gore area, and number of approaches at an intersection. This resulted in poor accuracy results for some asset types, including special pave- ment markings, roadside obstacles, rumble strips, sidewalks, intersections, lanes, and ramps. Other data elements were not collected at all, including pavement marking retroreflectivity, pavement edge drop-off, vertical curvature, sight distance, clear zone width, clear zone slope, superelevation, horizontal point of tangency, and location of traffic signal head. The research team provided some insight in the conclusions section of this report regarding why these data elements were not collected. On the basis of these observations, it is recommended that before commencing the data col- lection of the Safety Project S04B contract, the parameters and specifications should be clearly defined. As part of this effort, it is recommended that a Safety Data Dictionary and a detailed Field Data Collection Manual be developed to provide even more in-depth descriptions of data elements and to establish clear guidelines regarding where to reference the data elements, where to measure the data elements, how to process measured data into the desired format for future SHRP 2 work, and how to report parameters, including proper units. The manual should include photographs, diagrams, formulas, and any other guidance necessary to ensure that SHRP 2 receives the required data for Safety Project S04B. As part of the quality assurance process for S04B, it is recommended that short verification sites be established in each region to accommodate verification of the distance measuring instrument (DMI), GPS, image interval and quality, cross-slope, grade, and data take-off process from the images collected. The selected data collection contractor should be required to survey the veri- fication site prior to starting work in that region, as well as periodically throughout the regional data collection effort. It is recommended that SHRP 2 work closely with the selected team to ensure that it receives the data required to support the Safety Project S04, Acquisition of Road- way Information.

3The primary challenge faced during the S03 project occurred during the processing and evalu- ation of the data received from the participants. The challenges encountered with the participants’ data included the following: • None of the participants reported all 113 data elements requested. • The data provided were incomplete, and one team did not provide any final data. • Incomplete final data made it more difficult to link the data provided to the reference data set. • Very few teams provided geometric data. Some geometric data related to grade, cross-slope, hor- izontal point of curvature, length of curve, and radius of curvature were received; however, only two teams provided final geometric data in a usable format. These issues required more discussion with the participants to obtain additional data. Five of the 10 participating commercial firms were contacted to provide data to address these concerns. Two of these five firms stated that they would not provide any additional data (Teams 08 and 10). The other three firms (Teams 03, 06, and 09) provided additional data. However, the additional data were either in worse condition than the original (Team 06) or still incomplete (Teams 03 and 09). The research team spent additional resources cleaning the participants’ data files and linking the provided data back to their reference data sets. Because of the lack of geometric data received from the participants, SHRP 2 requested that participating firms reprocess their data to provide additional basic geometric and road inven- tory data. These basic geometric data included grade, cross-slope, horizontal point of curvature, length of curve, and radius of curve. The additional roadway inventory data requested included lane width, shoulder width (paved), and sign data (GPS coordinates and MUTCD code). These additional data are to be submitted to SHRP 2 along with the participants’ responses to the S04B request for quotations and proposals (RFQ/P) to be released in 2010. Despite the challenges with this project, S03 resulted in the following significant contributions to SHRP 2. • Research revealed several potential challenges and opportunities related to integrating road- way element data and in-vehicle data to be collected in S07, In-Vehicle Driving Behavior Field Study. These were documented in a white paper, dated October 10, 2008, and resulted in split- ting the S04 project into two components (S04A, Roadway Information Database Develop- ment and Technical Coordination and Quality Assurance of the Mobile Data Collection Project; and S04B, Mobile Data Collection). • In-depth insight and experience were gained regarding the practicality of collecting roadway safety elements. This resulted in the development of detailed recommendations for data col- lection practices to be used in S04B. • A list of data elements and associated specifications for collection during Safety Project S04B was generated.