U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
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|
Publication Number: FHWA-RD-01-051
Date: May 2001
Guidelines And Recommendations To Accommodate Older Drivers and Pedestrians
PDF Version (471 KB)
PDF files can be viewed with the Acrobat® Reader®
V. HIGHWAY–RAIL GRADE CROSSINGS (PASSIVE)
Background and Scope of Handbook Recommendations
According to the Federal Railroad Administration (FRA, 1999), in 1998, there were 3,508 highway–rail grade crossing crashes, resulting in 431 fatalities and 1,303 injuries. The majority of these incidents (64 percent) occurred during the day, 31 percent occurred at night, and 5 percent occurred during dusk/dawn. Fifty–five percent of the crashes in 1998 occurred at crossings with passive controls. In a National Transportation Safety Board study (NTSB, 1998), driver error was cited as the probable cause of the crash in 49 of 60 vehicle crashes analyzed at highway grade crossings with passive controls.
Klein, Morgan, and Weiner (1994) analyzed Fatal Analysis Reporting System (FARS) datafrom 1975 to 1992 to determine the characteristics of drivers involved in highway–rail grade crossings, and the circumstances under which such crashes occurred. This analysis indicated that drivers ages 25 to 34 are involved in the highest percentage (almost 25 percent) of all fatal rail crossing crashes, followed by drivers ages 16 to 20 (approximately 18 percent). Drivers in these age groups also show the highest involvement in all fatal crashes and all fatal intersection crashes, based on crash frequency data uncorrected for exposure. By contrast, drivers ages 65 to 74 were involved in 6.5 percent of fatal railroad crossing crashes and drivers ages 74 and older account for almost 5 percent of the railroad crossing fatalities. Again, these data do not reflect level of exposure. However, the data show that the percentage of drivers ages 65 to 74 who are involved in fatal rail crossing crashes is slightly more than the percentage of drivers in this age group who are involved in all fatal crashes (4.6 percent) and about the same as those involved in fatal intersection crashes (6.2 percent), which is the maneuver category for which seniors are most at risk. Notably, the proportion of older drivers involved in highway–rail grade crossing crashes at night is higher than the proportion of older drivers in vehicle–involved crashes at night, suggesting special problems associated with the use of these facilities under reduced visibility conditions.
There are several age–related diminished capabilities that may make the task of safely negotiating highway–rail grade crossings more difficult for older drivers. Well–documented losses in visual acuity and contrast sensitivity with advancing age (Burg, 1967; Ball and Owsley, 1991; Ball, Owsley, Sloane, Roenker, and Bruni, 1993; Decina and Staplin, 1993) may delay substantially the detection of critical elements such as the standard crossbuck or warning symbol during a motorist's approach to a crossing, and may preclude detection of a train actually present at the crossing until impact is imminent, especially at night. While the analyses of Klein et al. (1994) paint a compelling picture of young males engaging in intentionally risky behavior as a significant component of the crash problem at rail crossings, the technical literature suggests that willful noncompliance with traffic control devices by seniors at these sites will not be a major problem––if they (visually) detect and comprehend the advisory, warning, and regulatory information conveyed by these devices in time to respond safely.
Expectancy also plays a role in where and when drivers look for trains, and consequently, train detection (Raslear, 1995). A driver who is familiar with a crossing and rarely or never encounters a train during the time period he or she uses the crossing is more likely to miss seeing a train than either the driver who is unfamiliar with the crossing and therefore has no expectations about train frequency, or the driver who is familiar with the crossing and frequently encounters trains during the time period that he or she crosses the tracks. Drivers who don't expect trains do not look for them. As a consequence, per train, crash rates are higher for crossings with the lowest frequency of trains (Raslear, 1995). Enhancing the conspicuity and comprehension of design elements at passive crossings, plus the use of signing that orients drivers' attention toward trains and advises drivers on the appropriate action to be taken, are thus top priorities.
Comprehension of highway–rail crossing traffic control devices and performance of related information–processing tasks may be expected to pose disproportionate difficulty for older drivers. Although the crossbuck sign is a regulatory sign that serves as an implied YIELD sign, researchers consistently report that drivers do not understand the message it is intended to convey (Bridwell, Alicandri, Fischer, and Kloeppel, 1993; Fambro, Shull, Noyce, and Rahman, 1997).
Furthermore, assuming that a driver has been properly alerted to the need to search for an approaching train by design elements upstream and at the crossing, has slowed, and has begun to actively scan the tracks in each direction, the perception–reaction time (PRT) for a decision either to stop or to proceed, plus the subsequent execution of a brake or accelerator response, draw upon abilities found to slow significantly among the elderly (Staplin and Fisk, 1991; Goggin, Stelmach, and Amrhein, 1989; Stelmach, Goggin, and Amrhein, 1988). Whereas AASHTO (1994) uses a PRT of 2.5 s for calculating the sight triangle at passive grade crossings, over a decade ago, Gordon, McGee, and Hooper (1984) recommended that a full second be added to this design value to accommodate the 85th percentile driver. With the ever–increasing number and percentage of senior drivers, the need to refocus attention on this issue is urgent.
Additional insight is provided by Leibowitz (1985), who showed that inaccurate judgments of train speed and distance may make drivers' decisions to cross hazardous, due to perceptual illusions. Most drivers are not aware of the effects of the illusions of perspective, train size, and velocity (e.g., the bigger the object, the slower it appears to be moving), and this results in unsafe crossing decisions. Kinnan (1993) states that, in most cases, the driver believes the decision to cross is a rational one; most motorists seriously underestimate the risk because they can't properly gauge the speed of the train or its distance from the crossing. This problem will only be exacerbated by the age–related decline in the ability to integrate speed and distance information, as reported by Staplin, Lococo, and Sim (1993) for the judgment of gaps at intersections.
Finally, age–related hearing loss may contribute to a failure to detect a train approaching a crossing. According to government statistics (DHHS, 1994), approximately 30 to 35 percent of people ages 65 to 75 have a hearing loss, increasing to 40 percent for persons over the age of 75. Janke (1994) reported that totally deaf males have more crashes than their non–deaf counterparts, and drivers who wear hearing aids have an increased risk of crashing compared to drivers who do not wear them (excluding individuals who formerly wore hearing aids then discarded them, who had an even worse driving record). Thus, auditory train signals may not be completely effective as a secondary warning system for visually impaired drivers or drivers who neglect to properly scan at rail crossings if they are also hearing impaired. At the same time, data show that audible warnings can help reduce nighttime crashes, as evidenced by the 195 percent increase in collisions in Florida as a result of a nighttime whistle ban between 10:00 p.m. and 6:00 a.m. (Kinnan, 1993). Raslear's (1995) crash prediction model indicates that the use of the train whistle reduces the field of visual search from 180 to 10, which, in turn, reduces the visual search time by a factor of 18. By decreasing visual search time, the train whistle decreases the probability of a crash.
Though few studies have directly measured the effectiveness of countermeasures for older drivers in this arena, sufficient data exist to explain performance errors among the population at large to support highway–rail grade crossing design element recommendations for passive crossing control devices that offer the greatest promise to improve safety for older road users.