Do Drivers Stop at Unsignalized Intersections for Pedestrians Who Are Blind?
by Dona Sauerburger, COMS
March 25, 2003
Conference of Institute of Transportation Engineers, Ft. Lauderdale, FL
Literature from the Federal Highway Administration says that pedestrians should "always follow these steps when crossing a street:
1. Before crossing, stop at the curb, edge of the road, or corner before proceeding.
2. Look left-right-left and, if it's clear, begin crossing, looking over your shoulder for turning vehicles.
3. Continue to check for traffic while crossing."
Dozens of pedestrian safety brochures and web sites for children and adults give the same advice -- never assume that drivers will yield to pedestrians, and cross only when clear. The strong implication is that the pedestrian is responsible for avoiding collision. The Federal Highway Administration, for example, gives the following advice for crossing at traffic signals:
"Remember to make eye contact with drivers to ensure they see you. Don't take a walk signal, a green traffic light, or a driver for granted. Crossing safely is your responsibility. Remember, it's up to you." (Emphasis is from original text)
This caution is apparently warranted. In a recent study at two roundabouts in Annapolis, Maryland (Geruschat and Hassan, 2003), 10% of drivers failed to yield to pedestrians when entering one roundabout at an average of 15 mph, 32% failed to yield when entering the other roundabout with an average speed of 24 mph, 46% did not yield when exiting at an average speed of 16 mph, and 80% refused to yield when exiting at an average speed of 17 mph.
Are blind people exempt from this concern about drivers who fail to yield, and the implied responsibility of pedestrians to avoid collision? According to the law, yes. Every state has a variation of a law that requires drivers to stop or yield to pedestrians with white canes or dog guides.
However, every state also has laws requiring drivers to yield to pedestrians at crosswalks, and yet in many states we observe that drivers disregard this law. Do drivers obey laws to yield to blind pedestrians when they don't obey laws requiring them to yield to all pedestrians?
Some do, but most don't, according to my observations from more than 33 years of teaching people who are blind or visually impaired how to travel independently. For example, in 1999 I was one of 10 orientation and mobility specialists who went to the roundabout in Towson, Maryland to try to determine how blind people could cross there safely. We quickly discovered that we could not rely on drivers to stop for people using white canes. While standing on the median strip we attempted to cross the exiting lanes, but despite our white canes and the signs posted prominently reminding drivers to yield to pedestrians, very few drivers stopped. One of the drivers who failed to yield was driving a marked police car. I have had similar experiences when trying to cross separate right-turning lanes, and will show a video of 15 cars refusing to yield as they slowly passed me standing on the island tapping a white cane; they all stopped or slowed just after they had passed me so they could yield to oncoming traffic.
Research by Geruschat and Hassan (2005) at smaller roundabouts in Annapolis, Maryland verifies what we observed in Towson. When traveling at speeds averaging about 24 mph, almost two thirds of the drivers (63%) refused to yield to pedestrians who had walked up to the curb with a white cane and then stepped off the curb to place one foot in the crosswalk, placing the cane in the drivers' path. This was actually a great improvement over the rate of yielding to pedestrians without canes (only 10% yielded), but nevertheless, even when using a white cane, the pedestrian who relies on drivers to stop must accept considerable risk of a conflict or collision.
Since drivers will not reliably yield to pedestrians with white canes, blind pedestrians need to cross during sufficient gaps in traffic, just as other pedestrians do.
This is true at roundabouts, at separate right-turning lanes, and at intersections and crosswalks where there is no signal or stop sign for the street being crossed.
But unlike other pedestrians, blind people must determine gaps auditorially. They must be able to consistently and reliably hear the sound of approaching vehicles far enough to give them sufficient warning to allow them time to complete their crossing. For example, if they require 8 seconds to cross, they need to be able to hear even the quietest cars at least 8 seconds away, so that if they start their crossing just before they hear any cars (that is, while it is quiet), they will have 8 seconds to complete the crossing before cars arrive.
Whenever there are competing sounds, such as lawnmowers, air conditioners, wind or rustling leaves, and traffic noise, the ability to hear approaching vehicles can be severely reduced or even eliminated. For example, whenever a vehicle passes a blind pedestrian, the sound of that receding vehicle masks the sound of all approaching vehicles, making the pedestrian unable to hear any of them until they get too close. Therefore intersections must have intervals of complete quiet in order for blind people to detect all approaching vehicles sufficiently far to know that it's clear to cross.
However, there are many places where it's not possible to hear approaching traffic far enough even when it's quiet. In these situations, "all quiet" does not necessarily mean there is a sufficient gap to cross.
For example, at the roundabout in Towson, walking as fast as we could, we needed approximately 5 seconds to cross two lanes of traffic (curb to median). Thus, in order to recognize when there was a sufficient gap to cross, we needed to be able to hear cars at least 5 seconds away.
However, while standing at the curb when conditions were optimal (no masking sounds or wind -- it was "all quiet"), we were unable to hear many of the cars until they were only 2-3 seconds away. If we had stepped out just before we heard the cars approaching, we would have crossed in front of cars that were only 2-3 seconds from impact. Such a short warning would not provide enough time for us to clear the crossing before the cars would have reached us -- nor would it have been long enough for some drivers to be able to react and stop. When there was noise from traffic going around the roundabout it became even more difficult to hear many of the approaching cars. Often we did not detect them until they were in front of us, and sometimes we couldn't hear them at all.
When drivers do stop for blind people, masking sounds from other moving vehicles often make it impossible to detect the idling car, and so the blind person doesn't realize the car has stopped. When this happens, the driver will proceed after waiting a few moments for the pedestrian to cross. If there is more than one lane of approaching traffic and the blind pedestrian steps in front of a waiting car, he or she might be hit by a driver going around the waiting car in the next lane.
Blind people who cannot walk quickly or don't have good hearing or who take more time to process the gap information will be unable to safely cross some unsignalized intersections that are crossable by blind people who walk quickly, have good hearing, and can cognitively process various traffic sounds and make quick street-crossing decisions. Many of our elderly citizens have both impaired hearing and vision as well as a slower walking speed, and they will have difficulty crossing all unsignalized intersections safely.
In addition, cars are increasingly quiet, and hybrid electric cars are a popular choice in the dealers' showrooms. It won't be long before no one will be able to consistently hear all cars far enough to detect gaps reliably, even under optimal conditions with normal hearing and a fast walking speed. This will make intersections with uncontrolled lanes, such as separate right-turning lanes and roundabouts, inaccessible to visually impaired people who must rely on hearing to detect the gaps.
As we design our intersections of the future, we need to consider these issues. We should not wait until we find ourselves in a world where it's no longer possible to reliably detect gaps auditorially before we start to design intersections that are accessible to all pedestrians.
Geruschat, Duane R. And Hassan, Shirin E., [published in May (2005) "Driver Behavior in Yielding to Sighted and Blind Pedestrians at Roundabouts" Journal of Visual Impairment and Blindness, Volume 99 No. 5]
Survey of blind pedestrians: Their perception of yielding vehicles and ability to detect yields
Early in 2003, I sent a survey to several listservs of blind people, including several of the National Federation of the Blind and the American Council of the Blind, with the following message:
"Hi everyone! I'm trying to find out how consistently drivers in the U.S. stop for people crossing the streets with white canes or dog guides, because I'm going to do a presentation to traffic engineers at the Institute for Transportation Engineers later this month. I've done some experiments in California and will do some more in Maryland, and I want to see what's happening where YOU are!"
Fifteen people responded: The questions and responses were:
1. What state / city(s) do you usually travel in? Responses: They traveled in: Alabama (Auburn); Arkansas (Little Rock); California (Visalia, and San Carlos, near San Francisco); Louisiana (Houma); Michigan (Detroit); Missouri (St. Louis); Nebraska (Lincoln); Nevado (Reno); Texas (Austin, San Antonio); Utah (all over); Washington (Ballurse, Billingham, Burin, Olympia, Seattle); and Wisconson (Milwaukee)
2. Do you travel with a cane or a dog guide? Responses:
12 used a cane (several of these sometimes used no travel aid),
1 was a guide dog user, and
2 used both a cane and a dog.
3. When you are ready to cross, do you do anything to show drivers you intend to cross? (For example, do you stand at the curb or in the street, do you look a certain way, or move your cane or signal your dog, or what?) Responses: when using a cane:
"extend cane (About four or five times in the last ten years my cane was struck.)"
"stand on the curb facing the way I intend to cross."
"stand at the curb facing the street to be crossed."
"sometimes I signal to potential right-turning cars by holding out my left arm with my palm facing the car. Often, I wave the cane in front of me once just prior to stepping into the street"
"nothing out of the ordinary."
"stand at the curb with cane crossing front of body"
"hold my cane with the tip pointing towards the cars who will be turning on me so that they will see it. I always put one foot slightly forward showing my intent on crossing."
"look at the car and step out"
"step out just as any person might do. My cane is in front of me moving from side to side"
"move up to the very edge, look toward the vehicle, very briefly pause, then go, sending my cane out first, of course."
"swing my cane out and proceed across the street."
"stand at the curb and wave my support cane or long cane in the air."
"assertively ‘smack' my cane on the ground or make an exaggerated motion with my hand to indicate that I'm about to step off the curb."
"keep my cane tip up on the curb, then swing it out to begin walking across the street."
"Best rule of thumb is to assume no driver sees any pedestrian."
When using a dog:
"stand on the curb with a look on my face which clearly tells drivers that I know where I'm going."
"move my hand forward" [this hand movement is the command for the dog to start]
"stand with the dog at my side, waiting to cross, just as other pedestrians do."
4. Can you always tell when a car has stopped and is waiting for you? Responses: Only 3 out of 15 were confident and could always detect the yielding cars, and at least two of these 3 reported having vision (the other may also have vision, it wasn't asked).
5. In your state or city, if you travel in residential / suburban areas, about what percent of drivers yield for you after you show drivers that you intend to cross in the following situations? (If you don't normally cross in this situation, leave it blank)
Residential / suburban:
when you're at a crosswalk and there is no stop sign or signal for the cars on the street you're crossing (14 responses: average 64% yield);
when the drivers have a green light and are turning right across your path (14 responses: average 41% yield);
when you are crossing a separate right-turn lane from an island to the curb (or vice versa) (12 responses: average 41% yield).
In your state or city, if you travel in downtown or city areas, about what percent of drivers yield for you after you show drivers that you intend to cross in the following situations? (If you don't normally cross in this situation, leave it blank):
when you're at a crosswalk and there is no stop sign or signal for the cars on the street you're crossing (8 responses; average 51% yield);
when the drivers have a green light and are turning right across your path (14 responses; average 50% yield);
when you are crossing a separate right-turn lane from an island to the curb (or vice versa) (10 responses; average 41% yield).
Experiment to Determine
Rates of Yielding at Sites in California and Maryland for Pedestrian with White Cane at Separate Right-Turning Lane
Summary: At two sites in California, 15 out of 30 drivers refused to yield to a pedestrian who was standing in the street by the island, holding out a white cane and trying to cross the single channelized right-turning lane. At two sites in Maryland, 17 out of 40 drivers refused to yield. In all sites, most of those who yielded were patient and waited at least 10 seconds for the pedestrian to recognize they had stopped, and/or indicated they were waiting by honking, calling out the window, or getting out and guiding the pedestrian across.
In both situations, I stood with both feet in the street at a painted crosswalk, standing where the drivers could see me from a distance (in the second situation there was a clear line of sight a block away). I had the cane in front of me in the crosswalk, moving it in an arc side to side (similar to movement when walking).
Albany, California - Solano and Alameda Saturday February 15, 2003 partly cloudy 1:00-2:00 PM
Width of crosswalk = 17 feet
quarter circumference = about 39 feet
Yield? (Speed at 35 feet before crosswalk)
1. NO (Fast)
2. YES-honked (Slow)
3. Yes (Slow)
4. Yes (Slow bus)
5. Yes (Slow)
6. NO (Slow)
7. Yes (Slow)
8. Yes -- but proceeded after waiting only 3 seconds
9. Yes (Slow)
10. NO (Slow)
RESULTS: 3 out of 10 drivers refused to yield; 7 yielded.
Berkeley, California - Ashby and Adeline Monday February 17 2003 mostly sunny about 1:00-2:00
width of crosswalk = 24 feet
quarter circumference = about 41 feet
This time, I waited 10 seconds after they stopped, to see if they'd resume or wait and/or try to signal me that they had stopped
Yield? [If yes, waited/signaled?] (Speed at 35 feet before crosswalk)
1. NO (Fast)
2. Yes [Yes, waited 10 seconds] (Slow)
3. Yes* [NO -- after 2 seconds went through] (Slow)
4. Yes [Yes, waited 3 seconds and gave verbal go-ahead](Medium)
5. Yes [Yes, driver got out and assisted across](??)
6. Yes [Yes, waited 6 seconds and honked] (Slow)
7. NO (Moderate)
8. Yes [Yes, waited 4 seconds and gave verbal go-ahead] (Slow)
9. NO (Slow)
10. Yes [Yes, waited 8 seconds and gave verbal go-ahead] (Slow)
11. NO (Slow)
12. Yes [Yes, waited 1 second and gave verbal go-ahead] (??)
13. NO (Slow)
14. NO* (Slow)
15. NO (Slow)
16. NO (Moderate)
17. NO (Moderate)
18. NO (Moderate)
19. NO (Fast)
20. NO [after warning with honk](Moderate)
* Car immediately behind this car followed it and didn't stop. These cars were not counted in this survey.
RESULTS: 12 out of 20 drivers refused to yield. 8 drivers stopped to yield, all but one of these waited at least 10 seconds or indicated they were waiting by honking or calling out, or guided me across.
Both situations were well marked with pedestrian signs, one had faded crosswalk painted, the other had two white lines at the crosswalk. In both situations, I stood with both feet in the street standing where the drivers could see me from a distance (in the first situation the line of sight was about a half block away, in the second the line of sight was about a block long). I had the cane in front of me in the crosswalk, moving it in an arc side to side (similar to movement when walking).
Both times, I waited 10 seconds after they stopped, to see if they'd wait and try to signal me that they had stopped, or resume across the crosswalk.
Greenbelt, Maryland - Goodluck Road and Greenbelt Road Wednesday, March 12 sunny at first then partly cloudy about 2:30-3:30 PM
Width of crosswalk = 18.5 feet
quarter circumference = about 185 feet
Yield? [Waited/signaled?] (Speed at 35 feet from crosswalk)
1. Yes [yes, gave verbal go-ahead at 8 seconds] (Slow)
2. NO (Medium)
3. Yes [yes, honked at 2 seconds] (Slow)
4. Yes [yes, waited 20 seconds; I was unable to hear car] (Slow)
5. Yes [yes, waited 20 seconds] (Slow)
6. Yes [NO -- it was a motorcycle and after waiting 5 seconds, it kept going] (Slow)
7. NO (Medium)
8. Yes [yes, waited 20 seconds] (Slow-medium)
9. NO (Fast)
10. NO (Medium)
11. NO [this was a SCHOOL BUS!] (Medium)
12. Yes [yes, waited 20 seconds] (Slow)
13. Yes [yes, got out and guided across after 15 seconds] (Slow)
14. Yes [yes, gave verbal go-ahead after 11 seconds] (Slow)
15. Yes (Slow)
16. Yes [yes, it was a SCHOOL BUS and honked immediately when it stopped] (Slow)
17. Yes [yes, it was a SCHOOL BUS and after waiting 20 seconds it slowly drive past] (Slow)
18. NO (Slow)
19. Yes [yes, honked after 1 second] (Medium)
20. NO (Slow)
RESULTS: 7 out of 20 drivers refused to yield (one was a school bus driver). 13 drivers stopped to yield, all but one of these waited at least 10 seconds or indicated they were waiting by honking or calling out, or guided me across.
Wheaton, Maryland - Newport Mill and University Avenue: Wednesday, March 12 heavy overcast about 4:30-5:30 PM
Width of crosswalk = 18 feet
quarter circumference = about 72 feet
Yield? [Waited or signaled?] (Speed at 35 feet from crosswalk)
1. Yes [yes, waited 20 seconds] (Medium)
2. NO (Slow)
3. Yes [yes, waited 10 seconds] (Slow)
4. NO (Medium)
5. NO (Medium)
6. NO (Medium)
7. Yes [yes, honked after 1 second] (Medium)
8. NO (Medium)
9. NO (Slow)
10. NO (Medium)
11. NO (Medium)
12. Yes [yes, honked] (Slow)
13. Yes [yes, honked immediately] (Slow)
14. Yes [yes, waited 10 seconds] (Slow)
15. NO (Slow)
16. NO (Slow)
17. Yes [yes, gave verbal go-ahead after a few seconds] (Slow)
18. Yes (yes, drove around the island to avoid passing in front of me*) (Slow)
19. Yes [yes, gave verbal go-ahead after 7 seconds] (Medium)
20. NO (Slow-med)
* at least one other driver drove around the island to avoid passing in front of me but wasn't counted, so actually 10 of the first 20 drivers stopped or avoided me.
RESULTS: 10 of the first 20 drivers refused to yield. 8 drivers stopped and 2 drove around the island. ALL of the 8 who stopped waited at least 10 seconds or indicated they were waiting by honking or calling out, or guided me across or went around the island.