Bicycle Safety Guide and Countermeasure Selection System

 

Rectangular Rapid Flash Beacon (RRFB) at a Pinellas Trail Crossing

St. Petersburg, Florida

Prepared by Jake Thirsk, UNC Highway Safety Research Center

Background

St. Petersburg, Florida, had over 100 uncontrolled crosswalks, and in 2003, the city's pedestrian injury rate of 49.23 per 100,000 people was higher than both the county and state rates. In response, the city listed enhancements at uncontrolled crosswalks as a top priority in CityTrails, St. Petersburg's Bicycle Pedestrian Master Plan. Initially, RRFBs were installed at two uncontrolled crosswalk locations. The city studied and compared these RRFBs with other crossing enhancement devices and concluded that the RRFBs met with the greatest levels of success in terms of increasing motorist yielding rates. Therefore, they installed 17 more RRFBs and conducted a two-year review of their effectiveness. It was determined that system-wide average motorist yielding compliance rate improved from less than 1 percent to 82 percent and passes or attempted passes of vehicles that were stopped for pedestrians also decreased.

Given the results at the 19 uncontrolled pedestrian crosswalks, it was felt that the RRFB could also make a trail crossing safer by increasing the yielding of motorists to trail users, which includes bicyclists and pedestrians. In August 2008, an RRFB was installed at the Pinellas Trail crossing at 22nd Avenue, a busy, four-lane urban street. The "before" evaluation showed a motorist yielding rate of less than 2 percent and found that 18 percent of trail users were forced to stop in the middle of the intersection to wait for motor vehicle traffic. Traffic counts averaged at 15,000 vehicles per day with a posted speed limit of 40 mi/h. Trail users vary by season from 1,300 to 2,000 users per day. In the before and after period, over 80 percent of users were bicyclists.

Countermeasures

The push button activated RRFB included beacons and signs on the edge of the roadway and in the median. Each sign included two rectangular LED flashing beacons that were 6 inches wide, 2.5 inches high, and placed 9 inches apart. Both RRFBs were dual indicated (LEDs on front and back) and flashed 190 times in a 30-second, wig-wag sequence from left to right. Radio frequency transmitters linked each sign so that either of the push buttons activated both signs' flashers. A separate LED faced the trail user and flashed to indicate that the system was operating. Finally, an audible message indicated that the light across the street was functioning and reminded trail users to wait for cars to stop before attempting to cross.

The Pinellas Trail crossing in 2008.

The Pinellas Trail crossing in 2008.

Evaluation and Results

The Florida Department of Transportation commissioned a study to collect and analyze data on the crossing behavior of trail users and the stopping behavior of motorists both before and after the RRFB installation. Videotape data were collected for both directions of travel, in addition to supplemental observation data pertaining to pedestrian and bicycle counts, yielding compliance, and experienced conflicts. The main dependent variable that was coded from these data was whether the trail user or motorist yielded to the other. Yielding was defined as slowing or stopping. Where an avoidance maneuver or conflict occurred, the responses of the trail user and motorist were categorized.

Results were separated by trail user (pedestrian and bicyclist) and motorist actions, responses and interactions. The main purpose of the RRFB installation was to increase motorist yielding rates and reaction to trail users. The video tape data revealed that overall, motorist yielding increased from 2 percent to 35 percent. When the flasher was activated, motorist yielding increased to 54 percent.

Table 1: Yielding behavior before and after implementation of RRFB Table showing yielding behavior before and after implementation of RRFB.

Table 2: Yielding behavior depending on whether the button was pushed Table showing yielding behavior depending on whether the button was pushed.
Source: Tables 9 and 10 from Hunter, Srinivasan, and Martell (2008).

Trail users ability to cross all the way across the intersection increased from 82 percent before RRFB to 94 percent after. The supplemental data collected by trained observers also showed that overall trail user crossing delay was reduced.

Table 3: Bicyclists or pedestrians stranded in the crossing before and after the implementation of RRFB Table showing whether bicyclists or pedestrians were stranded in the crossing before and after the implementation of RRFB.
Source: Table 7 from Hunter, Srinivasan, and Martell (2008).

Conclusions and Recommendations

A definite and significant improvement in safety was noted by the fact that more trail users were able to completely cross the busy road without being stranded. While motorist yielding improved significantly after installation of the RRFB, an issue from the findings was why the motorist yielding was not greater. One factor is whether the button was pushed to activate the flashers.

According to the videotape observations after the installation, 32 percent of the trail users pushed the button, 49 percent did not, and for 19 percent of the trail users the button had already been pushed. The city installed a sign to notify trail users to push the button, but this issue remained a problem. Bicyclists tended to prefer to keep moving rather than stopping to push the button. General bicyclist behavior was to approach the crossing and slow to observe traffic and determine if a gap was suitable for crossing without activating the flasher. However, when traffic was heavy, bicyclists would stop to push the button and activate the flashers in an effort to expedite their crossing. Pedestrians were much more likely to push the button under all traffic conditions. It was also observed that motorists who were too close to the crossing when the flashers were activated tended not to yield. In many cases of heavy traffic, it took five to ten seconds after the flashers were activated before all the motorists would yield to the trail users waiting to cross.

Given these behavioral observations, researchers concluded that some additional educational efforts should be directed at increasing the percentage of trail users pushing the button and increasing motorists' awareness of their requirement to yield once the flashers are activated. Periodic police enforcement would assist these efforts. Additionally, the city is pursuing the development of a passive RRFB at the Pinellas Trail crossing where radar would be used to detect the approach of trail users desiring to cross and would automatically activate the flashers. However, such a system has problems that need to be satisfied before implementation (i.e., approaching bicyclists who can cross in periods of low traffic without need of the flashers, or pedestrians who stop to rest without the intention of crossing immediately).

Costs and Funding

The cost to furnish and install a flashing beacon can vary widely, depending on site conditions and the type of device used. The costs shown in the table include the complete system installation with labor and materials.

Infrastructure
Description
Median
Average
Min. Low
Max. High
Cost Unit
# of Sources (Observations)
Flashing Beacon
RRFB
$14,160
$22,250
$4,520
$52,310
Each
3(4)

The installation was completed on August 2, 2008 and included push button activation and beacons and signs on the edge of the roadway and median. The total cost of the system was $26,050.

Contact

Michael Frederick
Transportation Manager
St. Petersburg, FL 33712
(727) 893-7843

References

Bushell, M., B. Poole, C. Zegeer, and D. A. Rodriguez. Costs for Pedestrian and Bicycle Infrastructure Improvements: A Resource for Researchers, Engineers, Planners, and the General Public, p. 46. UNC Highway Safety Research Center. 2013.

Hunter, W., R. Srinivasan, and C. Martell. Evaluation of the Rectangular Rapid Flash Beacon at Pinellas Trail Crossing in St. Petersburg, Florida. No. FDOT Contract BA784. UNC Highway Safety Research Center. 2008.

Van Houten, R. and J.E.L. Malenfant. Efficacy of Rectangular-shaped Rapid Flash LED Beacons. n.d. Available: http://mutcd.fhwa.dot.gov/resources/interim_approval/ia11/stpetersburgrpt/stpetersburgrpt.pdf

Van Houten, R. and J. Shurbutt. Effects of Yellow Rectangular Rapid-Flashing Beacons on Yielding at Multilane Uncontrolled Crosswalks. Publication FHWA-HRT-10-043. FHWA. Washington, DC. September 2010. Available: http://www.fhwa.dot.gov/publications/research/safety/pedbike/10043/10043.pdf