Crossing Guard: Exploring Information Content in Navigation Aids for Visually Impaired Pededtrians
CHI 2012, May 2012, Austin, Texas, USA
Richard T. Guy
· Department of Computer Science, University of Toronto
· Ph.D. candidate
· He “explores the possibilities of computational imaging for assistive and collaborative interaction.”
· http://www.cs.toronto.edu/~guy/
Khai N. Truong
· Department of Computer Science, University of Toronto
· Associate Professor
· His “research interests are in human-computer interaction (HCI) and ubiquitous computing (Ubicomp).”
· http://khaitruong.com/index.phtml
Summary
The researchers are trying to solve the “problem of
orientation and mobility for visually impaired pedestrians by examining their
specific navigation needs.” Intersections are a point of stress, especially in
new areas. Feedback from visually impaired persons helped in the development of
CrossingGuard. The researchers performed a study to compare the information
given by CrossingGuard and in commercially available navigation tools.
Previous works have focused on orienting the impaired
person. A previous version allowed for dynamic rerouting around unexpected
delays. There are “several commercially available GPS navigation aids …
marketed directly to visually impaired people.” The designers need to know
exactly what information a visually impaired user needs for traversing an area
so that they can design effective tools. Recently, people have tried using
nonverbal audio and haptic feedback to provide information to impaired persons.
The ideas presented in this paper build upon some of the ones that have been
tested before.
The researchers held 2 structured interviews with 4 visually
impaired persons and 2 O&M specialists. “The first interview focused on the
strategies visually impaired pedestrians use to identify, navigate, and cross
both familiar and unfamiliar intersections.” The second interview “was focuses
on design criteria for a mobile navigation application.” Every participant stated
that they had trouble identifying alleys and driveways. Participants think that
“the device should include information about the shape of the intersection including
directions of available crossings, the width of streets, and the presence of
islands and other features of the roadway that affect crossing time” among
other things.
The CrossingGuard system provides “sidewalk to sidewalk”
directions for pedestrians with visual impairments. No novel ideas for
inputting the destination was explored due to the large numbers of previously
explored ideas. Users have simple gesture commands to support basic questions
such as “what is here?” A critical need for CrossingGuard is the interpreting
of data for use by visually impaired pedestrians. The prototype system was
based the OpenStreetMap project. The device describes each intersection based
on the number of streets in the intersection and the way in which they
intersect.
The researchers “designed a user study to test whether
having the more detailed information that CrossingGuard provides can raise the
comfort level of users as they prepare to cross an unfamiliar intersection
compared to base line information that is available on some commercially
available GPS navigation systems.” They
had 10 visually impaired people test the system, none of which had ever been in
the test area before. The participants were asked to describe the intersection
as they arrived and rate how comfortable they were with crossing the indicated
street on a 7 point Likert scale. After the walking was completed, the
participants were asked to participate in a brief follow-up interview.
The participants were asked “given the information you have
heard about this intersection, how comfortable do you feel crossing the street
here?” This showed that “having more
detailed information does increase comfort levels for visually impaired
pedestrians.” The participants were graded on the number of errors that were
committed along the route. According to the participants, “the shape, traffic
level, and traffic control device (light or stop) are the most salient features
of intersections.” Participants were asked to “rate the helpfulness of each
piece of information on a 7-point Likert scale.” The device pointed out when
alleys were present and most of the participants agreed with that feature. “A
common complaint was that intersections with very little traffic parallel to
the participant’s direction of travel were difficult to time.”
The researchers feel that “increasing comfort in new areas
is important because participants described feeling less comfortable traveling
to new areas or unfamiliar settings.” The participants offered suggestions for
people who are not advanced travelers. “Participants suggested that we extend
CrossingGuard to include the location of public transit stops and the location of
points of interest along the route.” Having the directions detailed by cardinal
directions confused some of the participants.
“To facilitate the collection of information that our
participants identified as most useful,” the designers developed a program
which would ask users to identify features of the intersection. Another way to
get additional information was the use of Mechanical Turk which has the users
answer questions about the intersection for a minimal payment. Mechanical Turk
users were able to easily answer some questions while others were more
consistently answered wrong. Figure 1 shows the kinds of information gathered by both of these applications.
 |
| Figure 1. |
“Navigation in new areas is a source of stress for visually
impaired people because they feel that they lack the information to travel
safely and confidently. The more information the participants had, the greater
their level of comfort.
Related Work
1.
“A New Approach for Pedestrian Navigation for Mobility
Impaired Users Based on Multimodal Annotation of Geographical Data” –
This paper focuses on a similar idea to CrossingGuard, the main difference is
that this research allows for the users to share data between them easily.
2.
“Pedestrian navigation aids: information requirements and
design implications” – This paper is all about the items that are
necessary to allow a person to navigate a new area; specific focus was not
given to those with visual impairments.
3.
“Cognitive Mapping and Wayfinding by Adults without Vision”
– This research is about the ways that visually impaired pedestrians create cognitive
maps of an area and how these skills are developed.
4.
“Understanding spatial concepts
at the geographic scale without the use of vision” – This article discusses the
way in which people with visual impairments understand geometric descriptions.
5.
“Navigation System for the Blind:
Auditory Display Modes and Guidance” – The idea behind this research is to
develop a “portable, self-contained system
that will allow visually impaired individuals to travel through familiar and
unfamiliar environments without the assistance of guides.”
6.
“From knowledge to words to wayfinding: Issues in the
production and comprehension of route directions” – This paper
discusses several of the problems associated with giving and understanding
directions.
7.
“Geography and the disabled: a survey with special
reference to vision impaired and blind populations” – This paper details some
of the basic problems that visually impaired people deal with daily and
provokes the reader to encourage the governments to make changes to help these
people get around in their daily life.
8.
“Exploring the Functional Specifications of a
Localized Wayfinding Verbal Aid for Blind Pedestrians: Simple and Structured
Urban Areas” – This paper proposes that a product highly similar to the one
discussed in CrossingGuard be created to increase the mobility of visually
impaired persons.
9.
“Non-Intrusive Somatosensory Navigation
Support for Blind Pedestrians” – This paper details a product that would allow
the visually impaired pedestrian to utilize their ears to hear what is around
them and keeps them on the correct path through the use of vibrators.
10.
“Verbal guidance rules for a localized wayfinding aid
intended for blind-pedestrians in urban areas” – This paper details
a few ideas for standardizing the information given to visually impaired
pedestrians.
As is clearly stated from the above listed papers, this idea is
not novel. There may be some parts which are more unique than others, but there
a large numbers of people trying to develop technology that will do exactly
what CrossingGuard does in the similar and different ways.Evaluation
There were two major forms of evaluation in this paper. The first was a qualitative, objective form
of evaluation on the number of errors made by the users. The second was a quantitative, subjective
form of evaluation; this was on the participant’s descriptions of the design. Each of these evaluation techniques was
effective, but I believe there could have been more numbers calculated and used
to evaluate the test. Overall, the users liked the material, but not everyone
agreed what was needed.
Discussion
I think that this paper covered a highly necessary piece of
research. Unfortunately, I do not
believe that the research done was entirely novel but the idea behind the
research is very helpful. As a human
being, I always feel better when I know the area where I am. I would imagine
that being blind would not change this for anyone. For that reason, I highly support the
research being done here.
No comments:
Post a Comment