Leading Pedestrian Interval --
A Solution We've Been Waiting For!
by Dona Sauerburger, COMS
with input from Michael King
All pedestrians who cross in the crosswalk during the green light have problems with drivers who turn into their path and refuse to yield to them. These drivers, who are disobeying the law, have been the cause of many pedestrian injuries and deaths, and they are one of the main concerns of blind people who are afraid to cross streets independently.
Now, there is a possible solution that reduces the danger for all pedestrians, called the leading pedestrian interval (LPI).
The LPI provides a few seconds head start to the pedestrians. The pedestrian signal comes on while the signal for the drivers remains red; after pedestrians have had a chance to start their crossing, the drivers get their green signal.
Fortunately, engineers in our area are starting to use LPIs. I noticed LPIs at intersections in Washington, DC over a year ago, and several months ago I noticed one for pedestrians crossing Wayne Street at Colesville Road in Silver Spring, Maryland.
LPI -- how it can work for blind people:
It is difficult or impossible for blind people to recognize when the intersection they are crossing has an LPI for them unless the intersection also has accessible pedestrian signals.
For example, several weeks ago I talked with two blind people who have crossed Wayne Street at Colesville for years.
They hadn't realized that the timing of the cycle had been changed and that the pedestrians were given a head start there. They thought that the pedestrians had become "brave" because so many of them started their crossing before the light changed to green.
Even when blind people realize there is a LPI, it is difficult for them to use it without an accessible pedestrian signal. I explained the LPI to my client who wanted to cross Wayne, but he was never able to recognize when the pedestrian signal came on until the traffic on Colesville had started.
By then, it was too late to start his crossing because of the aggressive right-turning drivers (the reason for installing the LPI there!).
However, if the signal can be provided with accessible pedestrian signals, blind people can take advantage of the LPI as well as sighted pedestrians, and reduce the danger and fear of crossing.
The rest of this article is excerpted from a paper, listserv messages, and report by Michael R. King on traffic calming devices installed at intersections in New York City. Michael is an architect and the former Director of Traffic Calming for the New York City Department of Transportation (NYC DOT) [NOTE written February, 2006: Michael King is currently with Nelson\Nygaard at 121 West 27 St, Suite 705, New York NY 10001-6207, phone 212.242.2490; His e-mail address is miking@trafficcalmer.com or mking@nelsonnygaard.com].
Michael will deliver his paper at the Urban Street Symposium in Dallas in June, 1999, and he graciously allowed us an advanced reading. This information can hopefully have a powerful positive influence on those who design our intersections, and may be useful to help us advocate for changes.
Do LPIs increase safety? by Michael R. King
The "pedestrian head start" is a sexier name for the leading pedestrian interval (LPI) or a delayed vehicle green. ["Pedestrian head start" is a play on words, as the "pedestrian head" is the pedestrian signal.]
Simply put, with an LPI, the WALK signal turns on 3 to 10 seconds before the vehicle green. This allows pedestrians to establish their presence in the crosswalk ahead of turning vehicles. The result is that drivers yield to pedestrians, instead of intimidating them by turning quickly.
LPIs have been used in New York City for at least 20 years. The NYC DOT has installed 55 LPIs at various intersections since 1975.
Recently we analyzed data and found that the LPI reduced the [crash rate between pedestrians and turning vehicles] by 12%. Relative to control sites, the rate fell 28%.1 Factored by severity, the relative rate fell 64% [see explanation in footnote #2]. Other studies have found similar results, and as a result, New York City is installing more LPIs, especially at locations of high turning movements.
In terms of potential safety benefits, we know that 35% of the 14,000 vehicle/pedestrian crashes in New York City every year involve turning vehicles. If the LPI truly reduces this number by 12%, then 600 vehicle/pedestrian crashes per year would be prevented.
These numbers are impressive, and considering that 2 out of every 3 people hit in crosswalks in New York City are hit while crossing with the signal, this bodes well for the future of the LPI as a tool to reduce pedestrian/vehicle crashes in New York City. Overall, we found that LPIs had the greatest effect on intersections with high turning movements, regardless of pedestrian traffic.
NYC DOT uses LPIs of 6-10 seconds. Richard Retting of the Insurance Institute for Highway Safety has studied LPIs with 3 second intervals and found similar results. [Ed. See below - article from WOMA newsletter 9/97.]
Analysis of Pedestrian/Vehicular Crashes at Leading Pedestrian Intervals in New York City (adapted from August 1998 NYC DOT report) by Michael King
Methodology
To understand how the LPI affects pedestrian safety, we analyzed fourteen locations with recently installed LPIs and substantial crash data. We obtained up to ten years' worth of crash data where possible (five years before and five after).
From this data, we were able to note vehicle action, pedestrian action, and the severity of the crash. The severity factor allows one to assign a numerical value based on cost to the public (ambulances, police, road repair, etc.) and then rate each crash accordingly [see note]. In order to assess the relative difference in crash rates, we collected data at the surrounding intersections. We also determined the statistical relevance of crashes at each site.
The basic analysis compares before/after data at the intersection to the before/ after data at the control site. This allows us to isolate the rate of change in crashes that would be affected by the installation of a LPI. These types of crashes include:
1. Total: all reported crashes involving a vehicle;
2. Crashes with an injured person(s);
3. Crashes due to driver error, such as failure to yield right-of-way, inattention, disregard for traffic control, turning improperly, etc.;
4. Crashes where a vehicle hits a pedestrian(s) in the crosswalk;
5. Crashes involving the pedestrian crossing with the signal;
6. Crashes involving vehicles turning left or right.
Theoretically the installation of an LPI will help to prevent and/or minimize all of the above crashes, especially types 3, 4 & 5. However, the LPI has a direct influence on type 6 -- those involving turning vehicles. We therefore used the statistics for this type of crash in assessing the performance of the LPI.
Findings:
Averaged together and compared to the control sites, the LPIs at the fourteen intersections yielded a 28% relative decrease in turning crashes, with a relative severity decrease of 64%.
In absolute terms [not comparing it to control intersections] , the LPI reduced the rate of crashes involving turning vehicles by 12%, with a 55% drop in severity [see note].
While the severity of crashes involving an injury rose 15%, and the difference in the rates of all reported crashes and crashes involving injuries was unchanged, the crash rates for the three other types of crashes affected by the LPI all fell from 27% to 88%.
Of the Manhattan locations with a significant decrease in LPI crashes, four have easily identifiable vehicle characteristics:
At East 14 Street, both University Place and Union Square West are one way into 14 Street. Therefore all vehicles must turn.
East 20 Street at First Avenue is the only through street between 14 and 23 Streets. Therefore most drivers wishing to access the intermediate blocks must turn at this intersection.
On East 62 Street, drivers wishing to access the Queensboro Bridge must turn onto 2 Avenue. Especially in the PM peak hours, this movement is quite heavy.
Vehicles traveling on East 96 Street to the 97 Street Central Park traverse must make a right turn on Madison Avenue, and then a left at East 97 Street.
Conclusions
The data shows that Leading Pedestrian Intervals have a positive effect on pedestrian safety, especially where there is a heavy concentration of turning vehicles. We surmise that this effect occurs regardless of pedestrian volume.
Discussion
There are about 11,000 traffic signals in New York City, about 85% have pedestrian indicators. A program to establish LPIs at all of these intersections would undoubtedly be a large undertaking. In that many of the signals are manually controlled, it would require breaking out and reconfiguring the cam. On the other hand, installing LPIs citywide, accompanied by a public outreach and education campaign, may save in the long run as fewer “wait for green†signs would be required, and the laws regarding yielding to pedestrians in the crosswalk would be accentuated from an engineering standpoint.
If taking 6 to 10 seconds from the green time has negative impacts on vehicle volume and level of service, one can trade the LPI seconds for increased steady DON'T WALK time at the end of the green phase. On streets that are 30 60 feet wide, the pedestrian (walking at 3 feet per second plus start up time) needs 15 25 seconds to cross. Given a typical 90 second cycle with 45 second phases, this allows for 20 30 seconds of time that can be dedicated solely to the driver. [A "phase" is the time in which one group or groups of drivers and pedestrians have their turn to cross the intersection; a "cycle" is the time in which everyone at the intersection gets the green signal once.].
NOTES:
1. Michael explained that crashes at intersections with LPIs fell an average of almost 12%, but during that same time period, crashes at the control sites increased almost 17%. Assuming that the crashes at the experimental intersections would also have increased 17% if they hadn't had LPIs, this means a reduction of 28% crashes (the difference between 12 and +17 is rounded out to 28).
2. Michael explained that crashes are factored in New York state by severity, based on the cost to society. The following factors that were used to rate the crashes were provided by New York's Department of Transportation:
Collisions with fatalities were multiplied by 2729, those whose victims were hospitalized and seriously injured were multiplied by 1214, those whose victims were hospitalized but not seriously injured were multiplied by 303, and those whose victims were injured but walked away were multiplied by 76. The total was then divided by the number of crashes.
The benefit in factoring crashes is that, while the number of crashes may not change much when an LPI is installed, if they are less severe, then overall we can judge that the location is safer.
In other words, it would be preferable to have 20 people with cuts and broken bones than 2 dead pedestrians.
In this case, even though the number of crashes was reduced by only 12% with LPIs, the severity of the crashes was reduced by 55%.
September 1997 newsletter, Metropolitan Washington Orientation and Mobility Association (WOMA)
Traffic Signal Adjustment May Make Street Crossing Safer by Dona Sauerburger, COMS
Many of our students have great difficulty crossing streets at intersections with traffic signals where there is frequent, heavy traffic turning into their path when the signal is green for them. The drivers often do not yield, and a number of visually impaired travelers in our area have been hit in these situations. Many of our students are understandably reluctant to cross at places like these, and some intersections are virtually impossible for blind people to cross because of this problem.
According to a publication of the Insurance Institute for Highway Safety that was written in the June 1997 "TRB Pedestrian Digest" ("Field Evaluation of a Leading Pedestrian Interval Signal Phase at Three Urban Intersections"), about 37 percent of pedestrian injury crashes and 20 percent of fatal crashes occur at intersections. It states that many conventional countermeasures try to address this problem by either encouraging pedestrians to notice potential "conflicts" with vehicles, or by encouraging drivers to yield to pedestrians. However, it will come as no surprise to most of us O&Mers that they report that "public education and enforcement campaigns have generally not produced tangible and long-lasting safety benefits."
Thus, a study was done to examine the influence of a three-second leading pedestrian interval (LPI). The LPI provides a brief, exclusive signal phase dedicated to pedestrians just before the traffic in the parallel street get their turn.
The result was that the use of a three-second LPI "reduced conflicts between pedestrians and turning vehicles, reduced the incidence of pedestrians yielding the right-of-way to turning vehicles, and appeared to make crossing the street somewhat easier."
In one study at a crossing with high volume of right-turning vehicles in Anaheim, CA (Hubbard, Bullock, and Thai 2008), the number of compromises (where the pedestrian had to delay or change path because of turning vehicles) dropped from 44 to 23 with LPI when the pedestrian was at the curb and from 6 to 2 while the pedestrian was in the crosswalk.
In another study at two interesections in Miami, Florida (Pecheux, Bauer, and McLeod, 2009), although the LPI did not change the percentage of yielding of right-turning vehicles, it significantly increased the yielding of drivers turning left into the crosswalk.
One study at 10 intersections at State Park, PA (Fayish and Gross, 2010) showed a 58.7% reduction in pedestrian/vehicle crashes with the LPI.
References:
Hubbard, S., D. Bullock, and J. Thai. Trial Implementation of a Leading Pedestrian Interval: Lessons Learned.
ITE Journal, Vol. 78, No. 10, 2008, pp 32, 37-41.
Pecheux, K., J. Bauer, and P. McLeod. Pedestrian Safety Engineering and ITS-Based Countermeasures Program for Reducing
Pedestrian Fatalies, Injury Conflicts, and Other Surrogate Measures Final System Impact Report. Federal Highway
Administration, U.S. Department of Transportation, 2009
Van Houten, R., A. R. Retting, C. M. Farmer, and J. Van Houten. Field Evaluation of a Leading Pedestrian
Interval Signal Phase at Three Urban Intersections. In Transportation Research Record: Journal of the
Transportation Research Board, No. 1734, Transportation Research Board of the National Academies,
Washington, D.C., 2000, pp. 86-92.