This research extends the results of laboratory research on wayfinding at intersections for pedestrians who are blind. Standard accessible pedestrian signals (standard APS), a prototype beaconing APS, and a raised guide strip were evaluated for their ability to assist in establishing and maintaining a heading for street crossings. Experiments were conducted at large, complex signalized intersections in Alpharetta, Georgia; Austin, Texas; and Towson, Maryland. Both the guide strip and the beaconing APS resulted in more accurate street crossing performance than standard APS with respect to alignment (i.e., initial heading) accuracy, rates of being within the crosswalk, distance from the center of the crosswalk at various points during crossing, and the likelihood of being well outside the crosswalk [6 ft (2 m) or more]. For the most part, performance with the guide strip or the beaconing APS was equivalent. Limitations and additional concerns with respect to these two treatments are discussed.

Five cues were evaluated with respect to their usefulness in directing the headings of pedestrians who were blind during street crossings. The study was conducted at a simulated crosswalk, with the angle of the crosswalk varied relative to the approach and direction of the slope of the ramp. Three cues worked well over the distance equivalent to the width of a six-lane road.

To travel independently, pedestrians with visual impairments must be able to locate crosswalks and align accurately for crossing streets. This action is especially challenging where crossings are not located at street corners and cues for locating the crosswalk and aligning to cross in the direction of travel on the crosswalk may be inadequate. This research compared midblock and roundabout crossings with curb ramps and detectable warnings but no other consistent cues for locating the crosswalk and no cues specifically intended to provide alignment and heading information, with the same crossings with the addition of a 24-in.-wide prototype surface of raised bars oriented perpendicular to the direction of travel on the crosswalks. This treatment was installed beside the detectable warnings and extended across the sidewalk. It was hypothesized that this surface would assist with both locating the crosswalk and establishing accurate alignment with the crosswalk. Sixteen participants having little or no vision failed, on average, to locate the crosswalk on the initial approach on 2.4% of their attempts for crosswalks having the guidance surface, while without the guidance surface, participants failed to locate the crosswalk on the initial approach on an average of 17.9% of their crossing attempts. With the guidance surface, participants aligned correctly on average on 77.3% of their trials, while without the surface the rate of correct alignment was 52.1%.

Pedestrians with vision impairments often have to cross streets at unfamiliar signalized intersections. The results are reported of research on crossings by blind pedestrians at complex intersections before and after the installation of accessible pedestrian signals (APS) and again following the installation of innovative device features. Objective data on measures of street crossing performance by 56 participants were obtained at four intersections, two each in Charlotte, North Carolina, and Portland, Oregon. The analysis includes broad measures of crossing timing, wayfinding, and independence. After APS installation, delay in starting to cross was reduced by approximately 2 s, and there was significant improvement in independently determining a safe time to cross, beginning to cross during the “Walk” signal, and completing crossings before the onset of perpendicular traffic. Some measures of wayfinding and related measures of independence also improved in Portland. Additional device features introduced after the first postinstallation testing session led to additional improvements in wayfinding during a second postinstallation test in Charlotte.

This study evaluates the usability of pedestrian signals by persons with varying visual acuities under different conditions of symbol size, crossing length and type of background behind the pedestrian signal. While viewing photographs presented on a computer monitor under unimpaired viewing conditions (approx. 20/20 visual acuity) and under simulated visual impairment (approx. acuities of 20/100 and 20/300), participants attempted to determine the pedestrian phase (Walk or Don’t Walk) and to report the number presented by the countdown timer display. Performance on the phase discrimination task by those with simulated 20/300 acuity and with 9 and 12 in. high symbols often resulted in performance which was little better than chance despite a highly controlled environment (i.e., no moving vehicles or environmental distractions) and signals which were subjectively in excellent working condition and of high visibility (i.e., good luminance/contrast, no glare). Reading the countdown display was essentially impossible. Participants with simulated 20/100 acuity were rather successful with regards to phase identification – averaging better than 87% correct under all stimulus conditions – but room for improvement exists as compared to performance in the 20/20 condition. Reading the countdown display was difficult for participants with simulated 20/100 acuity – averaging between 6.5% and 58.5% correct under the various stimulus conditions. The effect of different backgrounds on the usability of the signals, as well as the implications of the findings with regards to signal size and crossing length on the current signals standards, are discussed.