U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-16-035    Date:  June 2016
Publication Number: FHWA-HRT-16-035
Date: June 2016


Safety Evaluation of Intersection Conflict Warning Systems

Chapter 1. Introduction

Background on the ICWS Strategy

The ICWS strategy involves installing ICWSs on the approaches of rural, four-legged, unsignalized intersections. ICWSs may be installed on the major and/or minor approaches. These systems employ vehicle detectors to alert motorists of conflicting vehicles on an adjacent approach. Installation practices current at the time of this study used warning signs on the major approaches alerting motorists with the message “VEHICLE ENTERING WHEN FLASHING” (VEWF), “CROSSING TRAFFIC WHEN FLASHING,” or “WATCH FOR ENTERING TRAFFIC.” Signs on the minor approaches alerted entering motorists with “TRAFFIC APPROACHING WHEN FLASHING,” “LOOK FOR TRAFFIC” (with yellow light-emitting diode (LED) arrow-shaped flashers), or visual graphic displays. Figure 1 presents a Google Street ViewTM image of an ICWS application in Missouri. Refer to appendix A for further examples of ICWS applications observed in this study.

Figure 1. Photo. ICWS visual display from Google Street ViewTM. This photograph shows a post-mounted warning sign with flashing beacons above and below the sign face displaying the message stating "TRAFFIC APPROACHING WHEN FLASHING."

©Google® 2016

Figure 1. Photo. ICWS visual display from Google Street ViewTM.(3)

Use of ICWSs is one strategy employed at intersections with limited sight distance and/or intersections with a history of crashes involving gap acceptance problems. As Crowson and Jackels noted, there has been no specific guidance for the design, placement, and message of these systems, resulting in a broad range of approaches for States that are implementing these systems.(4) For this reason, the ENTERPRISE transportation pooled fund sponsored the research by Crowson and Jackels to develop a consistent approach for uniform deployment, provide further evaluation, and to recommend preliminary standards for the MUTCD.(2,4) Their research presented typical system components and developed recommended layouts for four scenarios based on which road the alert was directed and the number of lanes of the intersection. This research served as an evaluation of the safety effectiveness of ICWS applications to date through a crash-based analysis.(4)

Background on Study

In 1997, the American Association of State Highway and Transportation Officials Standing Committee on Highway Traffic Safety, with the assistance of FHWA, the National Highway Traffic Safety Administration (NHTSA), and the Transportation Research Board Committee on Transportation Safety Management, met with safety experts in the field of driver, vehicle, and highway issues from various organizations to develop a strategic plan for highway safety. These participants developed 22 key emphasis areas that affect highway safety.

The National Cooperative Highway Research Program published a series of guides to advance the implementation of countermeasures targeted to reduce crashes and injuries. Each guide addresses one of the emphasis areas and includes an introduction to the problem, a list of objectives for improving safety, and strategies for each objective. Each strategy is designated as proven, tried, or experimental. Many of the strategies discussed in these guides have not been rigorously evaluated; about 80 percent of the strategies are considered tried or experimental.

In 2005, to support the implementation of the guides, the FHWA organized a pooled fund study to evaluate low-cost safety strategies as part of this strategic highway safety effort. Over the years, the pooled fund has grown in size and now includes 40 States. The purpose of the pooled fund study is to evaluate the safety effectiveness of several tried and experimental, low-cost safety strategies through scientifically rigorous crash-based studies. The use of an ICWS was selected as a strategy to be evaluated as part of this effort.

Literature Review

Literature on ICWSs was limited. This section summarizes the salient research related to specific strategies. Very few studies were identified that investigated the effects of ICWSs.

Lyle evaluated a series of progressively more informative (and emphatic) signs used to warn drivers of a hazardous intersection at two locations in Maine.(5) The most informative (and emphatic) device was a warning sign stating “Vehicles Entering When Flashing” with corresponding flashing beacons. The measures of effectiveness for this study were observed speed reductions and driver sign recall. The active sign was found to produce the greatest decrease in speed, but the decrease was not significantly different from that produced by the next most progressive sign (“Vehicles Entering” with continuous flashing beacons). Surveys showed that motorists who saw the active warning sign had better recall, not only of the sign but also of the presence of a vehicle in the intersection.

Bretherton and Miao developed guidelines for traffic-actuated warning signs at intersections with limited sight distance based on data from 18 intersections in Gwinnett County, GA.(6) The 85thpercentile speed, existing sight distance, required minimum sight distance, and crash history were presented. The authors selected sites with at least three preventable crashes in 1 year, or at least one preventable crash for 3 consecutive years. A post-mounted “Vehicle Approaching” sign was used on the minor street approach for several intersections, and a post-mounted “Vehicle Entering Highway” sign was used on the major street approach for several intersections. The authors noted that “…the results show that the signs did effectively reduce the number of [preventable] accidents.”(6)(p. 12) Preventable crashes were defined as those related to limited sight distance. It should be noted that, because of the method of site selection, the results suffered from RTM bias.

Hanscom conducted a test of a collision countermeasure system in Prince William County, VA, using data from 1993 to 2000.(7) The primary measures of effectiveness were sign response speed, intersection arrival speed, first speed reduction, second speed reduction, and projected time to collision. Novelty speed effects were observed, but increased projected time to collision was sustained in the after period. A simple before–after crash analysis was conducted for side-impact crashes, and 2.6 crashes were observed per year for the 5-year before period, and no crashes were observed in the 2-year after installation period.

Peabody et al. also examined the effectiveness of a vehicle-activated warning system for stop-controlled intersections in Norridgewock, ME.(8) A conflict analysis showed a 35- to 40-percent reduction in intersection conflicts. A survey of drivers found that 67 percent said that the signs would prevent crashes, and 64 percent recommended the system for other intersections. Limited crash data were collected, and no crash effectiveness of the strategy was estimated.

The Pennsylvania Department of Transportation conducted a before–after analysis of a post-mounted collision avoidance system (CAS) at two locations from 1999 to 2005.(9) A speed study showed that operating speeds initially declined but increased after 3 years. A gap acceptance study found that typical gaps did not change from the before to the after period. Users were surveyed, and 97 percent said that the CAS was beneficial, and 93 percent said that the system should be installed at other locations. Summary statistics were presented for crash data. At onesite, two crashes were observed in the 2-year before period, and no crashes occurred in the 2‑year after period. At the second site, two crashes were observed in the before period, and threecrashes were observed in the after period. (One occurred while the system was malfunctioning.) The authors of the Pennsylvania study noted that the sample size was too small to conduct a safety analysis.(9)

The Missouri Department of Transportation (MoDOT) studied the safety effectiveness of post-mounted warning systems at 9 stop-controlled intersection major street approaches and 10stop-controlled intersection minor street approaches.(10) A simple before–after study found 28-, 72-, 37-, and 75-percent reductions in total, severe, angle, and severe-angle crashes, respectively, at the locations with the installation on the major street approach. They also found 32-, 33-, 44-, and 38-percent reductions in total, severe, angle, and severe-angle crashes at the locations, respectively, with the installations on the minor street approaches. MoDOT noted that one-third of the individual locations showed little or no improvement.

Simpson and Troy evaluated VEWF signs at 56 two-lane at two-lane intersections in North Carolina.(1) Installation dates ranged from 1996 to 2010. A before–after analysis assessed the crash reduction factor for multiple crash types. The following definitions were provided for the four categories of signs used in North Carolina:(1)

Table 1 presents the results of the analyses for two-lane at two-lane intersections. The authorsfound that deployments with alerts on the major road in advance of the intersection andlocationswith a combination of both major and minor road alerts were the most effective for two-lane at two-lane stop-controlled intersections, with CMFs for total crashes of 0.68 and 0.75,respectively.(1)

In addition, intersections with four lanes on the major route were considered; however, no apparent reductions in crashes were found for these sites. The authors suggested that VEWF systems may not be an appropriate strategy for most intersections with four lanes on the major route experiencing a strong frontal impact crash pattern.

Table 1. CMFs for VEWF signs.(1)
VEWF Category CMF Standard Error
Total Crashes
All sites 0.897 0.047
1 1.059 0.098
2 0.953 0.084
3 0.675 0.076
4 0.749 0.115
Target Crashes
All sites 0.929 0.055
1 1.074 0.112
2 1.001 0.096
3 0.679 0.088
4 0.797 0.144
Injury Crashes
All sites 0.878 0.059
1 0.917 0.108
2 0.934 0.106
3 0.732 0.102
4 0.870 0.187
Severe Injury Crashes
All sites 0.697 0.159
1 0.613 0.236
2 0.761 0.268
3 0.699 0.301
4 0.242 0.212

Statistically significant results at the 95-percent
confidence level are indicated in boldface.

Pierowicz et al. developed a prototype intersection collision avoidance system (ICAS) for use within vehicles.(11) The system was derived through the review of national databases such as the National Automotive Sampling System, General Estimates System, and Fatality Analysis Reporting System.(12) Four intersection crash scenarios were identified, as were three potential countermeasures. Two of the countermeasures, the Driver Advisory System and the Defensive System, were developed in a full-scale study for performance. Several recommendations were made from this research including the following:

Limitations of Previous Research

Most previous research studies focused on surrogate measures for intersection safety performance because typically only one or two applications were implemented. Some research studies were able to consider a simple before–after approach in an attempt to quantify a reduction in targeted crash types. However, these studies did not quantify a margin of error for the associated reductions, and they did not account for RTM bias. Only one of the studies reported in this chapter attempted to account for RTM, but details of how this was done were not provided. In addition, a linear assumption was used to account for changes in traffic volume experienced at the installation sites rather than safety performance functions (SPFs) typically used for EB evaluations. The study did find statistically significant crash reductions at the 95-percent confidence level for certain crash types for two-lane at two-lane intersections when all installation categories were combined. Not enough intersections and reference sites were available to study four-lane at two-lane intersections, and CMFs for several crash types for individual installation categories were statistically insignificant. Further, there was no attempt to quantify the impact of system placement on the major road (i.e., in advance of or at the intersection).



Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101