National Institute on Disability and Rehabilitation Research

Remote Infrared Signage Developmen to Address Current and Emerging Access Problems for Blind Individuals

Part I

Smith-Kettlewell Research on the Use of Talking Signs®
at Light Controlled Street Crossings

RESULTS

Performance of the Entire Group
The percentages of crossings that were completed successfully with and without Talking Signs were computed. The definition of "success" varied by measure. For example, in the measure of safety, participants either did or did not begin crossing within the walk interval. Additional data obtained regarding which part of the cycle they started in, were not considered in this analysis. If participants requested that the experimenter tell them when it was safe to cross, this was treated as missing data for the purposes of this measure of safety for that crossing.

In another example, in the measure of Precision "Did the participant end up within the crosswalk at the opposite corner?" participants either did or did not end up within the limits of the marked crosswalk at the opposite corner when they completed their crossing. If they required assistance for the crossing, this was treated as missing data for the purposes of this measure of precision. Additional data obtained regarding exactly where they did end up was not used for this analysis.

For measures of independence, success is equated with not requesting assistance or information and not requiring experimenter intervention for safety.

Table 1. Percent of successful crossings , by measure, with and without use of the Talking Signs system.

For the question regarding the shape of the intersection, participants were expected only to identify intersections as plus- or T-shaped. Additional descriptive information offered by participants was not utilized. For the question regarding the type of traffic control, participants were expected only to know that each intersection was controlled by a traffic signal. Additional information about the signal system was not used.

There were 80 crossings with Talking Signs and 80 without Talking Signs.

On every measure a higher percentage of trials was successful when completed using Talking Signs than without using Talking Signs.

Performance by subject, for each of the nine measures of street crossing success, with and without the use of the Talking Signs system was tabulated. Nineteen of 20 participants were more successful when using Talking Signs than when not using Talking Signs. One participant (# 11) had nearly flawless performance for all measures in crossings both with and without Talking Signs.

Table 2. Performance on 9 measures of street crossing effectiveness by each of 20 Ss with and without use of the Talking Signs system.

Binomial (step) tests were conducted in which each participant was compared with him/her self on each of the nine measures at each of the four intersections, in both the Talking Sign and no Talking Sign conditions, and then a statistic was computed to determine the probability that differences between performances using Talking Signs and no Talking Signs were significant. Each participant made four crossings using the Talking Signs system and four without. The data for safety indicate, for example, that for 14 participants more of these crossings were initiated during the walk interval when they were using the Talking Signs system than without, while for 6 participants performance was the same with and without Talking Signs. No participants were more successful in starting crossing within the walk interval without Talking Signs. (See Figure 4)

Figure 4. Binomial tests for 9 measures of street crossing performance with and without the use of the Talking Signs system.

Participants were significantly more successful (P<.05) on eight of the nine measures when using Talking Signs than when not using Talking Signs. The only measure in which there was not a significant difference in performance attributable to the use of Talking Signs was the need for assistance to complete a crossing.


Safety
Data were obtained on just when during the signal cycle street crossings were initiated. The standard procedure used by persons who are blind for crossing at signalized intersections is to start crossing when there is a surge of traffic parallel to the person's direction of travel. This surge of parallel traffic normally coincides with the onset of the walk interval. As all four intersections were signalized, participants should have initiated all crossings during the walk interval. At non-signalized intersections, the most common procedure is to cross when there is no traffic. Participants either thought they were starting their crossing during the walk interval or during a time when no traffic was present, depending on whether they thought the intersection was signalized or not.

Figure 5. Percentage of trials on which participants initiated their crossings on each interval of the signal cycle, in the Talking Signs and no Talking Signs conditions. (For no Talking Signs trials, N = 80, for Talking Signs trials, N=79 - one trial was omitted because the guide dog became distracted by the experimenters).

When using Talking Signs, participants started crossing during the walk interval on all but one trial, while without using Talking Signs, participants independently started crossing during the walk interval on only 53 of 80 trials and started during the flashing don't walk or wait intervals on 9 trials. When asked to cross streets without using the Talking Signs system, participants requested that the experimenter tell them when it was safe to cross on 18 trials (22.5% of crossings). At no time did participants request assistance knowing when it was safe to cross if they were using Talking Signs.

Choice of time to initiate a crossing is related to the type of control the blind pedestrian believes to exist at the intersection. On the nine trials without Talking Signs on which participants did not start during the walk interval, participants may not have believed the intersection was signalized. On five of these trials, participants started their crossing before a full cycle had elapsed, therefore it is likely that they did not think the intersection was signalized, and they were attempting to start during a gap in traffic. On three of the five trials on which participants started during WAIT, and one trial on which the participant started during the clearance interval, the participants were unaware that they were crossing at a signalized intersection.

Participants reported using eight different clues to determine when it was safe to cross. They were not limited in the number of clues they could report. Participants listed varying numbers of clues. It is likely that some participants used clues they did not report, therefore the list may not be fully reflective of the clues actually used by participants. In many cases both with and without Talking Signs, participants reported using more than one clue. In the Talking Signs condition, on a majority of crossings another clue was used in addition to Talking Signs information. A total of 133 clues were reported in the Talking Signs condition, 59 of which were not provided by Talking Signs messages. In the no Talking Signs condition, 79 clues were reported. The lower number for non-Talking Signs trials is likely accounted for by the 18 trials on which participants requested assistance because they judged that there was insufficient information to determine a safe crossing time, and because there was no single cue which was reliably available for all crossings.

Participants' self reports sometimes varied with experimenter observation. For example, one participant said he crossed when there was no traffic, but the experimenter observed a good surge of parallel traffic. Participants did not always report using Talking Signs when they were observed to have a Talking Signs "Walk sign" message playing just before starting a crossing. On only two crossings in the Talking Signs condition did a participant actually not have the "Walk sign" message playing before starting a crossing. On one of these, a bus pulled up to the destination boarding platform, perpendicular traffic was idling, and the participant started crossing without listening to the Talking Signs message. This was the one trial with Talking Signs on which a participant started during the WAIT interval.

Figure 6. Clues used to determine when it was safe to begin crossing.


Starting Location
Blind pedestrians typically prefer to start their crossings from within a marked crosswalk if there is one. Pedestrians are normally most visible to and anticipated by drivers when crossing within crosswalk lines. Participants were not asked to start within the crosswalk in this research, as it was desired to have the crossings be as representative as possible of crossings that would be made outside the experimental situation.

Figure 7. The percentage of trials on which participants started crossing from a position within the crosswalk. (N = 80 in the no Talking Signs condition; N = 79 in the Talking Signs condition because of missing data related to dog guide distraction by experimenters.

Pedestrians who are blind typically assume that when they come to an intersecting street, they are within or close to the crosswalk crossing the perpendicular street.

Some prefer to assume that wherever they arrived at a corner, they are within the crosswalk. If the sidewalks are narrow, this system often works well. If the sidewalks are wide, however, like those used in this research, blind pedestrians may not be within the crosswalk and may be some distance from the corner. In these situations, the blind pedestrian may search for a curb ramp because a curb ramp is required to be within the crosswalk. Dog guides will typically take their handlers to curb ramps.

In the no Talking Signs condition, on 14 trials participants were unable to locate the curb ramp or other information which made them confident that they had found a good starting location and they requested experimenter assistance. Participants were much more successful at starting within the crosswalk in the Talking Signs trials, however, where they would not get the message about the status of the pedestrian signal unless they were within the crosswalk and facing toward the destination curb.

Alignment
On 101 of the 159 total trials, participants began their crossings from a heading toward the destination curb, aligned so their projected path of travel would be within the crosswalk. On 54 of these 159 trials the heading was not toward the destination curb. In the no Talking Signs condition, heading toward the destination curb was at chance level.

Figure 8. The number and percentage of trials on which participants were aligned so that they were headed toward their destination curb, in Talking Signs and no Talking Signs conditions. (N = 80 in the no Talking Signs condition; N = 79 in the Talking Signs condition because of missing data related to dog guide distraction by experimenters.)

Participants reported using eleven different clues to align themselves in the direction of the destination curb. They were not limited in the number of clues they could report. Participants listed varying numbers of clues. It is likely that some participants used clues they did not report, therefore the list may not be fully reflective of the clues actually used by participants. In many cases both with and without Talking Signs, participants reported using more than one clue. In the Talking Signs condition, on a majority of crossings another clue was used in addition to Talking Signs information. A total of 146 clues were reported in the Talking Signs condition, 61 of which were not provided by Talking Signs messages. In the no Talking Signs condition, 94 clues were reported.

In the Talking Signs condition participants described their use of the system to establish a heading toward the up curb by facing the direction in which they received the clearest signal and/or based on the Auxiliary message description such as "The crosswalk is angled away from Fulton." Participants in the no Talking Signs condition reported using information from the travel patterns of other pedestrians and information provided by moving perpendicular traffic on more crossings than in the Talking Signs condition.

Figure 9. Clues used to establish alignment toward the destination curb.

Ending Location
Ideally pedestrians who are blind would prefer to complete street crossings within the crosswalk at their destination curb. To this end, they attempt to start crossing from within the crosswalk and to establish a heading toward the destination curb, a heading which would normally be within the crosswalk. While veering to some degree is common while crossing streets, there are a number of clues that pedestrians who are blind can use while crossing, to help them walk straight toward the destination curb. These clues include the presence of parallel traffic, other pedestrians crossing the street, idling cars on the perpendicular street, and the crowning of the street.

On one crossing (Grove Street at Polk Street) participants in the no Talking Signs condition were often aligned so that they faced somewhat into the intersection. However, beyond the middle of the street the crowning was so pronounced that participants were guided toward the destination curb by the crowning. Almost half (11 of 20) of the trials in which participants started crossing misaligned, were completed within the crosswalk.

Blind pedestrians who do not complete their crossings within the crosswalk may veer into the intersection and then correct by veering back toward the destination corner because of any of the clues to walking straight, or because they realize they have traveled too far and reason that they must have veered toward the middle of the intersection. They may then correct their direction and complete their crossing near the destination, but a little way up the parallel street from the cross walk. Alternatively, they may veer slightly away from the intersection and complete their crossings near the destination corner along the perpendicular street, but outside the cross walk. These are fairly common problems, not usually resulting in great danger or disorientation to the blind pedestrian.

Occasionally a pedestrian who is blind will head toward or veer into the middle of an intersection. On these occasions the pedestrian may continue across the intersection to the diagonally opposite corner, sometimes crossing a parallel street on which vehicles have a green light. Alternatively, they may end up on the corner across the parallel street from the corner on which they started. These crossings expose the pedestrian to considerable risk and may be very disorienting . There were two crossings of these types when using Talking Signs and four when not using Talking Signs.

On 16 (of 40) crossings at Market Street, where participants were asked to find a boarding platform instead of crossing the entire width of the street, participants missed the platform, sometimes crossing the entire street width and assuming that they had only then arrived at the platform. This occurred an equal number of times on Talking Signs and no Talking Signs conditions. Apparently the descriptive, Auxiliary message (information for finding the platform) was not sufficiently precise or well understood. There was no Talking Signs transmitter on the platform, and all information came from the vicinity of the ped head directly across the street. Therefore the Talking Signs user had only the description message for guidance to the platform.

Figure 10. The percentage of trials on which participants ended their crossings in various locations relative to the crosswalk. N = 80 in the no Talking Signs condition; N = 78 in the Talking Signs condition due to dog guide distraction and technical difficulties.)

On 8 (of 20) crossings of High Street at Fulton Street (T intersection) without Talking Signs, participants veered into Fulton Street (the stem of the T) far enough to encounter the line of parked cars along Fulton Street. Some were quite disoriented. No participant veered far enough into the parallel street to encounter the line of parked cars when they were using Talking Signs. At this intersection, Talking Signs users had both the direction of the message providing pedestrian interval information and the descriptive information about the direction of the crosswalk, to enable them to accurately head for the destination curb.

Individual Differences
Four participants reported mild to moderate hearing loss. They showed essentially the same pattern of results as the total group of participants. They may have been slightly more careful and successful travelers. Performance was equal to or better than that of non-hearing impaired participants on safety, precision, independence and knowledge.

Figure 11. Performance of four participants with mild to moderate hearing loss on safety, precision, independence and knowledge when using Talking Signs and when not using Talking Signs for street crossings, compared with performance of the other 16 participants.

Two participants do not normally travel or cross streets in unfamiliar areas and don't consider themselves to be good or well oriented travelers. They were very successful when using Talking Signs, but had quite limited success when crossing without Talking Signs.

Figure 12. Performance of two participants who do not normally cross at unfamiliar intersections, and who report that they are fair travelers with limited orientation skills, when using Talking Signs and when not using Talking Signs for street crossings, compared with performance of the other 18 participants.


Subjective Assessment
Following the experimental procedure, participants were asked a number of questions to provide a subjective assessment of the technology and the need for training in its use. Participants first responded to seven statements by rating the extent of their agreement on a five point scale in which 1 = strongly disagree, and 5 = strongly agree.

Table 2. Subjective rating of usefulness of the Talking Signs system at intersections. Number of participants who rated each statement at each level. 1 and 2, indicating disagreement, are grouped, and 4 and 5 indicating agreement, are grouped.