Data Collection Practices for Use with In-Service Performance Evaluations (2024)

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.

2 BACKGROUND Generally, the performance of new roadside hardware such as longitudinal barriers, sign supports, guardrail terminals, impact attenuators, and work zone devices is evaluated initially using the currently adopted full-scale crash testing procedures [3, 4, 7-9]. The American Association of State Highway and Transportation Officials (AASHTO) Roadside Design Guide (RDG) provides guidance on how to select, locate, and place roadside hardware [1]. The crash test evaluation criteria are limited by their very nature to a small number of impact speeds and angles, specific vehicles, properly installed hardware, and laboratory conditions. After successful crash test performance has been demonstrated, the roadside hardware is installed in the field according to the policies of the appropriate transportation agency. The AASHTO RDG is a reference available for transportation agencies when developing their own regionally appropriate policies. Ideally, the field performance could be assessed by the transportation agencies using an ISPE as described in NCHRP Research Report 1010: In-Service Performance Evaluation: Guidelines for the Assembly and Analysis of Data or previously using the procedures provided through NCHRP Report 490: In-Service Performance of Traffic Barriers [5, 10]. ISPE is the process of examining in-service collisions to determine how effectively roadside hardware works when exposed to the full spectrum of vehicle types and impact conditions as well as the site, maintenance, weather, and traffic conditions that the roadside hardware is exposed to in the field. During hardware development, particularly the crash testing phase, there is no question that the hardware was properly constructed and in optimal crash-ready condition. The effect of prior crashes, variations in site design, maintenance, or improper construction cannot be assessed in a full-scale crash test because the tested hardware was in ideal condition. Crash testing does not guarantee satisfactory performance in the field under all conditions over the life of the hardware. ISPEs of roadside hardware and features have been performed sporadically since the mid- 1970s [10]. These early ISPEs were generally one-time efforts performed to answer a specific performance question. Many of these early studies would be categorized today as routine or retroactive studies since the data used consisted of already collected law enforcement reports, maintenance data, and other transportation agency data. Fortuniewicz and Bryden appear to be the first researchers to extend their data collection activities into the field in their study of turned- down end treatment for heavy-post median barriers in the State of New York in 1983 [10]. 2.1 ISPE DATA COLLECTION PROCEDURES There are very few examples of formalized ISPE data collection procedures. The Federal Highway Administration (FHWA) recently completed a Pilot ISPE of guardrail end terminals (GET) which collected performance data in five transportation agencies: CALTRANS, MoDOT, PennDOT, MassDOT and the Pennsylvania Turnpike Commission [12]. The experiences and lessons learned by these five transportation agencies were documented and developed into a model data collection procedure. The data collection procedure discusses developing and using data collection forms and taking photographs of the crash scene, damaged hardware, and, if available, the damaged vehicle. NCHRP Report 490 included a detailed field data collection procedure and data element dictionary in Section 4 (Data Collection - Site Investigations) of Appendix D [13]. NCHRP 3 

Report 490 includes a list of equipment needed, techniques for making measurements, guidelines for taking photographs, and example data collection forms. 2.2 ISPE DATA ANALYSIS GUIDELINES NCHRP Report 490 includes a discussion of analyzing ISPE data and provides instructions for calculating crash rates, risk probabilities, and comparing systems in Section 7 (Analysis) of Appendix D [13]. More recently, the Transportation Research Board (TRB) initiated a policy review study to examine the issue of using ISPEs to study the performance of guardrail end terminal (GET) hardware. The team was given the task of reviewing the literature, interviewing experts, and assessing the feasibility of performing ISPE of GET hardware. The results were published in TRB Special Report 323: In-Service Performance Evaluation of Guardrail End Treatments. TRB Special Report 323 concludes by saying “state highway agencies will require more information about the benefits, costs, and practicality of routine in-service evaluation of end treatments in general before deciding to undertake new data collection and analysis programs necessary to carry out more challenging analyses.” The committee recommended further research to advance ISPEs and test the feasibility of and costs associated with more complex evaluations. It also recommended research to examine whether procedures for testing the performance of devices should be altered. Most recently, NCHRP Project 22-33 developed “Guidelines for the Assembly and Analysis of ISPE Data,” which was published as NCHRP Research Report 1010 [2]. NCHRP Research Report 1010 identified the minimum data elements and sample size needed for an ISPE of any roadside safety feature. NCHRP Research Report 1010 outlines a two-tiered approach to conducting an ISPE. The first tier, a routine ISPE, maximizes the use of available, already collected data through integration of existing transportation agency data sources (e.g., crash data, maintenance records, insurance recovery records, etc.). The second tier, an investigative ISPE, is directed by the results of the assessment in the first-tier routine ISPE. If the first-tier routine ISPE results in acceptable performance, then no additional data collection is needed and the second-tier investigative ISPE is not undertaken. On the other hand, if it is found in the first-tier routine ISPE that additional data elements are needed to form conclusions, then the second-tier investigative can be initiated. The ISPE analyst uses the routine ISPE results to determine the data collection area, time frame, and the data elements to collect in the field for the more detailed investigative ISPE. One obvious advantage to routinely collecting ISPE data elements is the second, more resource intensive part of the NCHRP Research Report 1010 methodology, will not be triggered. Not all transportation agencies have the resources for collecting every data element on every type of roadside hardware, therefore, a focused two-tier approach is advantageous and efficient. 2.3 ISPE EVALUATION MEASURES NCHRP Research Report 1010 detailed the standardized evaluations measures for assessing hardware performance under real-world conditions. Evaluation measures are stand- alone. Each evaluation measure uses standardized data elements and standardized analysis techniques to facilitate coordinated ISPE practices across the states. Consideration is given to the safety feature under evaluations (SFUE). This identifier is intended to be a broad category (e.g., longitudinal barriers, terminals, and crash cushions, etc.). An SFUE is identified by a number which represents the broad grouping of safety features, as shown here: 1. Longitudinal Barriers 4 

2. Terminals and Crash Cushions 3. Truck- and Trailer-Mounted Attenuators and Variable Message Signs and Arrow Board Trailers 4. Support Structures, Work Zone Traffic Control Devices, and Breakaway Utility Poles 5. Other Features Note that specific information about each of these broad groups is not provided here. The declared SFUE establishes which ISPE evaluation measures are appropriate for consideration. The evaluation measures assess the following performance measures: (1) the structural adequacy of the safety feature under evaluation; (2) occupant risk through consideration of the crash severity; and (3) post impact vehicle trajectory through consideration of the crash sequence of events [2]. Each NCHRP Research Report 1010 evaluation measure is discussed here. 2.3.1 Evaluation Measures for Structural Adequacy 2.3.1.1 Evaluation A (Safety Feature Breach) Longitudinal barriers (i.e., SFUE=1) are installed to reduce the instances of vehicles which cross from the traffic side to the field side of the barrier. This evaluation measure is intended to evaluate the containment and redirection abilities of longitudinal barriers and is therefore limited to SFUE 1 systems. 2.3.1.2 Evaluation B (Breakaway) Support structures, work zone traffic control devices, and breakaway utility poles (i.e., SFUE=4) are designed to readily activate in a predictable manner by breaking away, fracturing, or yielding. This evaluation measure is intended to evaluate the breakaway, fracture, or yielding abilities of support structures, work zone traffic control devices, and breakaway utility poles and therefore is limited to SFUE 4 devices. 2.3.1.3 Evaluation C (Controlled Penetration, Redirection, Stopping) Terminals and crash cushions (i.e., SFUE = 2) and truck- and trailer-mounted attenuators and variable message signs and arrow board trailers (i.e., SFUE = 3) are generally designed to allow for redirection, controlled penetration, or controlled stopping of the impacting vehicle. This evaluation measure is intended to evaluate the redirection, controlled penetration, or controlled stopping abilities of terminals and crash cushions and is therefore limited to SFUE 2 or 3 devices. 2.3.2 Evaluation Measures for Occupant Risk 2.3.2.1 Evaluation D (Occupant Compartment Penetration) Evaluation D is intended to evaluate the influence of and propensity for occupant compartment penetration on crash severity across the vehicle and speed mix the safety feature is exposed to while in-service. 2.3.2.2 Evaluation F (Rollover) Evaluation F is intended to evaluate the influence of and propensity for rollover that results from interaction with the safety feature under evaluation. 2.3.2.3 Evaluation H (Vehicle Mix) Evaluation H is intended to evaluate occupant risk across and within the vehicle and speed mix the safety feature is exposed to while in-service. Evaluation H assesses the crash severity in terms of the maximum injury experienced by the impacting vehicle’s occupants. 5 

2.3.3 Evaluation Measures to Assess Vehicle Trajectory 2.3.3.1 Evaluation J (Secondary Impact on Roadside) Evaluation J is intended to evaluate the relative risk of post impact secondary crashes with fixed objects to those crashes with no secondary impact with roadside fixed objects. 2.3.3.2 Evaluation K (Secondary Impact on Road) Evaluation K is intended to evaluate the relative risk of post impact secondary impacts on the roadway compared to those crashes with no post impact secondary impact on the roadway. 2.3.3.3 Evaluation L (Impact Orientation) Evaluation L is intended to evaluate the relative risk due to the impact orientation. Evaluation L considers SFUE 2 devices (i.e., terminals and crash cushions), SFUE 3 devices (i.e., truck- and trailer-mounted attenuators and variable message signs and arrow board trailers), or SFUE 4 devices (i.e., support structures, work zone traffic control devices, and breakaway utility poles). 2.3.3.4 Evaluation M (Impact Orientation) Evaluation M is intended to evaluate the relative risk due to the impact orientation for SFUE 1 devices (i.e., longitudinal barriers). 6