Design Considerations and Evaluation Methodology for Adapted

Navigational Assistants for People with Cognitive Disabilities

Javier Gomez and Germ´an Montoro

Department of Computer Engineering, Universidad Aut´onoma de Madrid, Madrid, Spain

Keywords:

Assistive Technologies, Mobile Navigation, Interface Design Recommendation, Evaluation Methodology,

Cognitive Disabilities.

Abstract:

Assisting pedestrians with cognitive disabilities in their movements through a city is not a simple task. Despite

of the new mobile navigational software available in the market, many of these users (and their caregivers) are

still resistant to use them. In some cases, due to the lack of adaptation to their needs. This issue motivated us

to elaborate a set of design considerations to keep in mind when designing navigational assistants for people

with cognitive disabilities. Besides, we developed a navigational prototype for smartphones that was evaluated

with two users with Down syndrome. Therefore, we also propose some hints about how the evaluation should

be carried out.

1 INTRODUCTION

How can I get from my position to some other place?

This question can be answered thanks to our mobile

phone and a navigational software. However, it could

be more difﬁcult if the user had some kind of dis-

ability. In this paper we focus on locating, giving

directions and helping users with cognitive disabili-

ties, and more speciﬁcally Down syndrome, in their

walking movements through the city. By means of

the smartphone as a location and information presen-

ter device, we want to promote their independence,

autonomy and self-conﬁdence, besides giving support

to caregivers.

The particular necessities of these people make

even more difﬁcult to understand the maps or direc-

tions that current navigational software provides. Not

only their cognitive condition may limit them to un-

derstand complex instructions, but they usually also

have some other disabilities, e.g. dyslexia, aphasia,

vision problems, etc. that should also be considered

when designing any assistive resource.

Current navigational software provides adaptation

mechanisms, such as different palettes for day or

night light, adding oral interaction or even adapting

the route depending on the means of transportation

(car, walking, public transport or bike) or the trafﬁc

status in real time. Nevertheless, they do not take into

consideration the different cognitive load required for

users with special necessities, such as Down syn-

drome people.

In order to address the design and evaluation

processes of the assistive navigational software, we

present a set of considerations based on the litera-

ture and the experimental research. These could be

useful to ease the development and evaluations of fu-

ture approaches. These considerations have been put

into practice to design and develop an assistive navi-

gational prototype for pedestrians with cognitive dis-

abilities. The prototype has been evaluated with two

users with Down syndrome in a real environment.

This paper is organized as follows: after review-

ing the backgroundof this topic and the current frame,

we extract a set of design considerations. Then, we

describe the prototype of a mobile client. Next, we

present a case study, and ﬁnally we propose some

design considerations and suggest a methodology for

evaluating this kind of assistive software.

2 BACKGROUND

In the recent years, the interest of the research com-

munity on technologies and services speciﬁcally de-

signed and developed to help people with special

needs has grown. Not only caused by human or per-

sonal factors, but also supported by governments and

institutions. Evidence of this is the ISO 9999 (ISO,

2011) that deﬁnes Assistive Technologies (AT in for-

ward) that are “any device, equipment, instrument

344

Gomez J. and Montoro G..

Design Considerations and Evaluation Methodology for Adapted Navigational Assistants for People with Cognitive Disabilities.

DOI: 10.5220/0005203603440351

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2015), pages 344-351

ISBN: 978-989-758-068-0

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

or software produced to prevent, compensate, moni-

tor, calm or neutralize disabilities in the body struc-

tures or their functionalities, restrictions in activities

or social participation”. Or the Assistive Technol-

ogy Act of 1998 (ATA/1998) (105th Congress of the

United States, 1998) which also deﬁnes the AT as

“product, device, or equipment, whether acquired

commercially, modiﬁed or customized, that is used to

maintain, increase, or improve the functional capabil-

ities of individuals with disabilities”.

Both deﬁnitions are very similar but they bring a

common framework that can be assessed from differ-

ent faces. In general, giving assistance depends on the

user necessities and capabilities. Thus, the problem

faced in this case study, giving navigational support,

as an assistive technology artifact depends directly

on the user’s disability as well. For example, when

calculating a route and giving directions to a person

with motor disabilities (e.g. she needs a wheelchair to

move), it would more interesting to suggest an acces-

sible route, without stairs or narrow pavements, than

the fastest path. On the other hand, in the case of

people with cognitive disabilities it would be prefer-

able to adapt the route in terms of cognitive load, i.e.

reduce the number of decision points and their com-

plexity. Therefore, before studying, designing and de-

veloping any assistive artifact, user necessities should

be stated.

Thus, in this case study we focused on giving di-

rections to people with cognitive disabilities. These

disabilities may come from birth or acquired later,

for example, brain injury after a stroke. Moreover,

depending on the clinical proﬁle, necessities may be

different. Therefore, in order to provide a precise vi-

sion of the problem, we restrict this study to address

the issues concerning to give directions to people with

Down syndrome in their walking movements through

the city.

3 LITERATURE REVIEW

The state of the art offers a considerable number

of studies, developments and open questions regard-

ing different disabilities. However, when focusing

on users with cognitive disabilities they reduce dras-

tically. The complexity of the brain and the dif-

ferences between subjects make this research area

difﬁcult but challenging at the same time. Brad-

dock et al. evidenced this issue in (Braddock et al.,

2004): ATs have been traditionally designed for peo-

ple with different disabilities, such as motor control

(robotic prostheses) or communication (augmented

and alternative communication devices). But they

rarely studied or developed to assist people with cog-

nitive disabilities. Despite of this hurdle, literature

reviews still provide interesting projects and stud-

ies. In particular, many of them are focused on

assisting people with cognitive disabilities in their

daily lives, such as GUIDE (O’Neill and Gillespie,

2008), COACH (Mihailidis et al., 2000), Autonoma-

mente (Barbieri et al., 2010), ePAD (Mihailidis et al.,

2010), ATHENA (Hidalgo et al., 2011), or 2D-Tasks

(Caballero-Hern´andez et al., 2012).

Particularly, we focus on navigational assistants,

which are rarely prepared for people with cognitive

disabilities. In this sense, mobile application mar-

kets offer a signiﬁcant number of navigational sys-

tems, supported by big technological companies e.g.

Google, Apple, Microsoft or some other specialized

in navigation and maps, e.g. Tom Tom, Garmin, etc.

Besides, almost all of them provide some adaptation

mechanisms. Regarding the interface they usually of-

fer different languages, daylight/night mode, 2D or

3D maps, etc. But also the route calculation is usually

adapted to the means of transportation or the trafﬁc.

However, in most cases the cognitive load is not con-

sidered when adapting both the interface and the inter-

action. A map with an icon to represent the user and a

highlighted route is the most common view. Around

this map a lot of extra information is included, such as

speed, the next instruction, GPS signal quality, time

to destination, etc. These screens, full of informa-

tion, continuously updated and, in many cases, not so

easy to understand make these applications not suit-

able for guiding people with special needs. Besides,

according to Boisvert et al. (Boisvert et al., 2009),

when designing mobile aids for people with cognitive

disabilities, the cognitive load should be reduced, by

means of offering customized experience, multimodal

information, giving regular feedback and preventing

errors. Related to interface design, some important

notes are given in (Cole, 2011). In the paper the au-

thor introduces the Patient Centered Design approach

that integrates User-Centered Design and Participa-

tory Design to develop assistive systems adapted to

patients necessities (in this case users with cognitive

disabilities).

Anyway, there are some interesting research pa-

pers in this topic such as MAPS, by S. Carmien

(Carmien, 2006) (Carmien, 2003), which is a prompt-

ing PDA software to assist people with special needs.

Moreover, if it’s combined with a GPS coordinates

server, it can be programmed to guide the user in out-

doors movements. Another interesting research is the

one done by Richter (Richter et al., 2010), that em-

ployed user models to improve the user experience

and (Richter and Duckham, 2008), a discussion on

DesignConsiderationsandEvaluationMethodologyforAdaptedNavigationalAssistantsforPeoplewithCognitive

Disabilities

345

how instructions should be given. Finally, Liu et al.

presented in (Liu et al., 2009) the design process and

the evaluation of a guiding system, exploring the suit-

ability of visual widgets and directions for users with

cognitive disabilities. On the other hand, authors such

as Beeharee and Steed (Beeharee and Steed, 2006),

Fickas et al. (Fickas et al., 2008), Liu et al. (Liu

et al., 2009), Lemoncello et al. (Lemoncello et al.,

2010) explored different mechanisms to give direc-

tions to people with cognitive impairments, namely

using arrows, audio prompts, maps and landmarks. In

general, all authors agreed that the use of landmarks

is the best way to orient and guide these particular

users. Additionally, Garcia et al. (Garc´ıa de Marina

et al., 2012) explain some other issues, such as the

difﬁcult to distinguish between left and right (when

using textual instructions) and the confusion that pro-

duced the difference between the actual view and the

picture shown, i.e. elements that appear or do not ap-

pear.

Therefore this is a challenging topic, moreover if

we consider that, according to (Carmien, 2010) peo-

ple with cognitive disabilities usually present limita-

tions on spatial navigation skills that makes more dif-

ﬁcult to guide them.

4 DESIGN RECOMMENDATIONS

As it was stated in the previous section, current navi-

gational applications do not present adaptation mech-

anisms for cognitive impaired people. However, some

previous studies offered interesting ideas, such as

the use of landmark-based instructions, which pro-

duced better results in the experiments. On the other

hand, simple and clean interfaces are usually valu-

able guidelines when designing applications for peo-

ple with cognitive disabilities. Additionally, studies

like (Kr¨uger et al., 2004) or (Ishikawa et al., 2008)

stated that, in some cases, the assistive device could

act as a distractor. Therefore, the design of the inter-

face is a key task.

Besides, it is very important to take into consider-

ation the necessities of the caregivers. Providing as-

sistance with an adapted system should decrease the

effort necessary to do it by traditional methods. Adap-

tation of the system to the environment and the users

should facilitate this task.

Considering these previous works and relevant

ideas extracted from the researched literature, and

with the assistance of experts and caregivers, we pro-

pose a set of considerations that may help when devel-

oping assistive navigational software for people with

cognitive disabilities:

• The less possible information on the screen, the

better. Interfaces should be clean and simple.

Users should only receive relevantinformation for

their goal. For instance, for an advanced user it

would be interesting to know the subsequent de-

cision points. For cognitive disabled people, this

information could be confusing or distract them

of their goal.

• Destination points should be selected from a pre-

viously created list. Regular uses, in many occa-

sions, employ their navigation systems to go to

new and unknown places. Nevertheless, people

with cognitive disabilities, and more speciﬁcally

Down syndrome, do not usually move alone to

new places. The places they are usually allowed to

go by themselves are restricted to a certain num-

ber. This way, the destinations should be con-

strained to a previously set up list. Additionally,

this also relieves the user to look for places (us-

ing searching boxes) or remembering additional

information about the destination (e.g. exact ad-

dress).

• Real images of the landmarks should be combined

with atomic instructions. Studies like (Richter

and Duckham, 2008) demonstrated that the per-

formance is related to the way that directions are

given. In this sense, atomic instructions i.e. “con-

tinue straight” or “turn left” in combination with

landmark identiﬁcation ﬁt better with the user ne-

cessities and help them in the learning process.

This way, showing an image of the landmark and

telling the user to “continue straight until you see

this” may lead to a better user experience and an

improvement in the learning process. Addition-

ally, the cognitive load is reduced since these in-

structions are easier to understand than combined

instructions, such as “turn the second on the left

and then right”.

• A navigation system should promote their auton-

omy. The most immediate task of these systems

is to guide users to the destination but, addition-

ally, they should promote their independence. To

employ such systems should not only provide the

knowledge for a concrete route, but also to help

users to learn new routes and landmarks identiﬁ-

cation. Since users have to pay attention to real

environment landmarks that makes them familiar

with their surroundings and what they ﬁnd in their

routes.

• The system should provide additional information

to assist users in the process to identify the land-

marks. This can be done by notifying them about

the distance to the next decision point or land-

HEALTHINF2015-InternationalConferenceonHealthInformatics

346

mark. This information can be presented in dif-

ferent ways, but it should consider how to reduce

the cognitive load. Therefore, instead of show-

ing measures like meters or a scale, the system

may be provided of a coloured ﬁlling bar. This

is an easy-to-understand and friendly interface.

Besides, using colors provides another source of

feedback, since variations in color attract user at-

tention effectively (Brown, 1999). Additionally,

as Garc´ıa–Herranz et al. demonstrated in (Garc´ıa-

Herranz et al., 2012), people with cognitive dis-

abilities usually relate colors to states in the same

way as users without disabilities. Green means

good, red bad and amber in the middle. This way,

a green bar that ﬁlls as the user walks or turns in

the correct point provides useful and simple infor-

mation.

• The system should have mechanisms to prevent

and solve errors. People with cognitive disabili-

ties present lower error acceptance so prevention

should be dealt by means of awareness. The sys-

tem should be aware of users’ position and dis-

tance, and notify them when they are reaching

a decision point or if they are taking a wrong

direction. On the other hand, getting or feel-

ing lost may produce a stressful situation. Keep-

calm messages or communication channels with

the caregiver should be considered for these situ-

ations.

• Caregivers and relatives should be kept in mind.

Therefore, communication and supervision tools

should be provided. Particularly, it would be in-

teresting for the caregiver to be able to locate the

user at any time, or study the routes followed to

analyze her performance.

5 MOBILE NAVIGATION

ASSISTANT PROTOTYPE

Before developing a new system focused on a speciﬁc

group of users it is very important to be sure of how

you should face their problems and design the inter-

face they are going to use. This is how, with the de-

sign considerations presented in the previous section,

we designed a ﬁrst prototype to validate the function-

ality of such a system. The employed architecture is

irrelevant for this study and only a general description

of it is presented.

In a nutshell, thanks to the current on-line services

(such as Google Maps) that provides tools to model

environments, calculate routes and counts with huge

street level pictures databases, we were able to de-

crease the time caregivers have to spend to model new

environments and take photographs of the landmarks.

In order to process and adapt all this available infor-

mation to our mobile client, a middle server was de-

veloped.

Once the user chooses the destination point (from

a previously loaded list, stored in the server), the

server requests the map service for the route between

the two points, split it, and asks for the images associ-

ated to each landmark. Finally, all this is information

is sent to the user’s phone in a proper format.

5.1 Mobile Client

The mobile application is the most important element

of this study. From the user’s point of view it is the

module that she has to deal with. So, considering

the user’s special needs, the graphical interface and

means of interaction have to be carefully designed.

Therefore, following the previously stated ideas we

decided to develop the interface shown in Figure 1.

As can be seen, the interface is divided into different

sections (from top to bottom):

Figure 1: Interface screenshot.

The application name is in white font over dark

grey background. These colors were intentionally se-

lected to make it clear that this is not an active part of

the interface. Under it, the instruction is displayed

in black font over light green background. Apart

from the text, the instruction is read aloud as the in-

terface loads. Then, we ﬁnd a red button with the

text “Help”. Again, colors are intentionally selected,

meaning “something is wrong”. This button provides

the prevention and error correction mechanisms. This

is, if the user feels lost or in panic, she can press

the button and the caregiver would be automatically

notiﬁed about the stress situation and the location of

the user. After that, in the middle of the interface,

as the most important element, the system presents

the real picture of the landmark that the user has to

ﬁnd. In this case it is a street-level view extracted

DesignConsiderationsandEvaluationMethodologyforAdaptedNavigationalAssistantsforPeoplewithCognitive

Disabilities

347

from Google Street View Service. Right under the im-

age, the distance bar is showed. It is slightly ﬁlled of

green, which means that the user is still “far” from the

decision point and, therefore, from that image. This

feedback mechanism helps the user to be aware of her

performance. Additionally, once the bar gets almost

full, the phone rings and vibrates to notify the user

that the decision point is close. Finally, two naviga-

tion buttons are placed at the bottom part of the in-

terface. Regarding the design we did not change any

color or shape, and placed them as regular buttons.

The one on the left, with the text “Previous” is used in

case the user wants to go back and repeat from the last

decision point to make sure of the route. The other

one, “Next”, is pressed when the user reaches a de-

cision point and wants to receive the next instruction

and its associated landmark picture.

As it was said in previous sections, instructions

have to be simple and easy to understand. This is why

these systems should split the route into atomic steps.

For this case study prototype every decision point is

divided in two parts: ﬁrst, the user is asked to con-

tinue straight until she ﬁnds the picture shown in the

interface. Once she is at that decision point, she is

asked to turn left or right and, then, a new street-level

picture of the next landmark is shown. Finally, she is

asked again to continue straight until she reaches the

decision point showed in the picture.

In order to illustrate it, in Figure 2 we sketched an

example route. In this case the user has to go from

the post station (blue circle with a white envelope) to

the bank (green circle with a white dollar symbol in-

side). The proposed route follows the red numbered

dots (from 1, starting point, to 5, destination point).

As it can be seen, the other red dots correspond to

decision points. Besides, for every decision point we

present two pictures: the ﬁrst one to provide the user

with a view of the landmark and, the second one, to

help her in the turns. Pictures are adapted to the turn-

ing direction. This is, they are showed according to

the point of view of the user.

Figure 2: Route sketch to go from the post station to the

bank. The decision points has been highlighted with num-

bered red dots.

6 CASE STUDY

In order to evaluate our navigation assistance concept

and the proposed design considerations, we ran a ﬁrst

case study experiment with two participants. They

both were Down syndrome youngmen. As they stated

in the interviews, they were not used to going alone to

near places, although their relativeswanted them to do

it. This prototype tried to ﬁt their needs.

The experiments took place in their home town

and users were required to walk a route of 600 meters

between two main landmarks: the train station and

the City Hall. To reach the destination, they had to

perform two 90 degrees turns to the left. As they did

not live in the city center, the route was completely

new for them.

6.1 Methodology Proposal

To evaluate the system it is important to deﬁne a pre-

cise methodology. For this study, and following the

literature, we decided to divide the experiment in dif-

ferent sessions and tests. In ﬁrst place, both users

were asked to complete two questionnaires. The ﬁrst

one about personal information, technology and inter-

net habits. The second one to evaluate their self-report

about navigational tasks.

This way, with the ﬁrst test, we were able to get

an initial idea of their capabilities, needs and relation

with technology. Besides, they also had a ﬁrst train-

ing session with the prototype, so that they could get

familiar with the system.

Then, they were asked to answer the Santa Bar-

bara Sense Of Direction Scale (SBSODS) test by

Hegarty et al. (Hegarty et al., 2002), in order to get

their self-report on how they considered themselves

about giving and understanding directions or maps.

This test consists of 15 questions to be answered in a

1 to 7 Likert scale (1 strongly agree, 7 strongly dis-

agree). Nevertheless, in order to provide a reliable

and neutral test, we proposed 8 questions written in

a positive way (e.g. “I am very good at giving di-

rections”) and another 7 questions in a more negative

sense (e.g. “I have trouble understanding directions”).

Therefore, positive answers scores should be inverted

so that higher values can correspond to better results.

After that, in a new session, participants were

asked to walk through the proposed route, using a

phone equipped with the navigation assistant proto-

type. In order to gather ﬁrst-person thoughts, the

“think aloud” procedure (Lewis, 1982) was carried

out to get extra subjective information about user’s

feelings and opinions. Besides, according to (Lepist¨o

and Ovaska, 2004) these techniques provide more in-

HEALTHINF2015-InternationalConferenceonHealthInformatics

348

formation than usability tests or questionnaires. To

do that, each participant was asked to wear a head-

mounted camera and a voice recorder.

Finally, after the walk, participants were asked to

take two additional tests and a short interview. The

ﬁrst test consisted of a sequence of real images of the

decision points. With them, they were asked about

the direction they had taken. This is a widespread

mechanism to measure the spatial knowledge (Siegel

and White, 1975). After that, the USE Questionnaire

(Lund, 2001) was given to the users to evaluate the us-

ability of the system and ask them their opinion. This

test provides user’s opinion of four topics, namely

“Usefulness”, “Ease of use”, “Ease of Learning” and

“Satisfaction”. Finally, a short oral interview was car-

ried out to ask them about the things they liked most

and least.

6.2 Results

The experiments yielded interesting results to deﬁne

and complete the previously presented design consid-

erations and methodology for evaluation:

• First of all, we checked that the guidance ap-

proach worked: splitting a complex direction in

atomic instructions and presenting street level im-

ages to identify landmarks made users to ﬁnd the

route easily and arrive at their destination.

• Moreover, thanks to the progress bar they were

able to estimate the remaining distance to the next

decision point without effort. In fact, according

to the subsequent interviews, the progress bar was

the most valuable part of the application, followed

by the real images.

• Regarding to user’s performance we can highlight

that both users reached the destination in an ac-

ceptable time (less than tree times the calculated

time) without any external assistance besides the

navigational software.

• According to the later study of the recordings,

their performance was slightly different, but they

shared some common attitudes. For example,

both participants paid more attention to the phone,

looking for the landmark, as they got closer to it.

This is important since it demonstrates the impor-

tance of the use of alarms to help them to identify

the landmarks.

• In contrast, the required time to identify land-

marks and directions at the decision points was

different. Participant 1 needed less time on av-

erage. When asked about landmarks identiﬁca-

tion, none of the users answered right to all the

questions, but again their responses were differ-

ent. This shows how important it is that the sys-

tem adapts to each speciﬁc user as well as how

they need to repeat the same route multiple times

in order to get conﬁdent with its steps.

• On the other hand, according to the interviews,

the users and their caregivers, found the prototype

very interesting. They were eager to use them

again in future routes as a way to ﬁnd their way

without the need of external help.

7 DISCUSSION

7.1 Design Considerations

From the carried out evaluation, we conﬁrmed and ex-

tended the design considerations that should be fol-

lowed in the process of developing mobile assistive

navigational software for people with cognitive dis-

abilities. The following are the characteristics that a

navigation assistant should satisfy in order to ﬁt these

users needs:

Regarding the interface design:

• Users and caregivers should be included in the de-

sign process. This way we ensure that we design a

prototype to solve problems without creating new

ones. By including experts (i.e. therapists, educa-

tors, tutors, etc.) in the design process we may

avoid future errors and develop the appropriate

adapted software.

• It should be simple and clean, avoiding distract-

ing information. To do that, the design should be

friendly, iconographic and minimalist in order to

catch user’s attention.

• Multimodality should be supported (texts, images

and voice) in order to overcome reading or vi-

sion problems. Many users present some other

disabilities, such as low vision, which should be

considered while designing the interface. In or-

der to overcome some of these other disabilities,

providing the information in different ways could

allow the application to be used by a wider range

of users.

• Colors usually have the same meaning for users

with special necessities that for users without dis-

abilities. Nevertheless, and related to the previous

point, color blindness should be taken into consid-

eration to chose a proper palette.

• Feedback should be provided, but not acting as a

distractor. Users need to know whether they are

doing right or not, but in a non-intrusive way that

DesignConsiderationsandEvaluationMethodologyforAdaptedNavigationalAssistantsforPeoplewithCognitive

Disabilities

349

does not require full attention.Besides, the sys-

tem should be aware of user’s progression and be

proactive, notifying whenever it is needed.

• Due to the lower error acceptance of these users,

prevention mechanisms should be included.

Regarding the route calculation and guidance:

• In order to save human and time resources, space

modeling, route calculation and landmark images

should be as automatically processed as possible.

• The use of images to identify landmarks is ac-

cepted as an accurate way for guiding. The more

realistic and updated the images are, the easier

will be to identify landmarks.

• Atomic instructions (continue straight or turn

left/right) seem to work properly. Therefore, to

divideroutes into these instructions is a good tech-

nique.

• To provide continue and accurate feedback con-

tributes directly to the user’s performance.

7.2 Methodology for Evaluation

In section 6.1, we proposed a methodology for evalu-

ating these systems. After the case study, we noticed

some aspects that we could highlight:

• Participants should not be overwhelmed. Putting

the user in unpleasant or uncomfortable situations

without a reason may affect her opinion or per-

formance. Asking and listening to caregivers and

even given them the opportunity to participate

may become the environment more familiar. This

will facilitate that the user can interact and partic-

ipate actively.

• Performance relies on users’ capabilities. There-

fore, the more we know about them, the better. In

other words, personal data or habits are valuable

information that may provide a ﬁrst idea of the

possible user performance. However, large tests

or interviews may make users get tired, affecting

their mood, answers or performance.

• In general, Likert scale tests do not work well with

these users. Their answers are usually biased to

the extremes so they should be avoided. Instead

of them, more open questions, which let users ex-

plain their thoughts will be better. Furthermore,

the “think aloud” method has demonstrated to be

particularly useful when evaluating systems with

cognitive disabled people.

• Although SBSODS is widely known as a good

measurement of users’ self-opinion about naviga-

tional tasks, they should be adapted for cognitive

disabled people.

• Recording users’ point of view provides valu-

able information, both qualitative and quantita-

tive. Head mounted cameras or camera-equipped

glasses are affordable choices.

• Time and location registries are another source of

analysis of user’s performance. The more precise

is the gathered information, the deeper the studies

will be able to be.

8 CONCLUSIONS

In this paper we have presented a set of design rec-

ommendations to take into consideration for the de-

velopment of new adapted navigational assistants for

people with cognitive disabilities. Besides we pro-

pose a methodology as long as recommendations for

evaluating these systems. In order to validate these

ideas, we developed a navigational prototype to assist

pedestrians with cognitive disabilities in their walking

movements through the city.

This prototype was tested with two users with

Down syndrome in an open environment. Since the

number of tests is not big enough to validate the sys-

tem it becomes a ﬁrst case study that can support the

initial ideas presented in this paper as long as offer

new considerations to take into account.

The design guidelines were based, at the begin-

ning, in literature review. But, thanks to the ﬁeld trial

and the experience gathered with these ﬁrst users, we

have been able to have an initial conﬁrmation and ex-

tension of our proposals. The same happened with the

evaluation methodology.

Since this is a ﬁrst case study we consider that it

can already provide some useful considerations about

the design and evaluation processes in the develop-

ment of adapted navigational assistants for people

with cognitive disabilities, as stated in previous sec-

tions. Therefore, for future work the prototype should

be tested with more users, re–adapting the interface,

design guidelines and evaluationmethodologiesto the

results obtained with a wider variety of users and sit-

uations.

ACKNOWLEDGMENTS

This work has been partially funded by “Fundaci´on

Vodafone” and “Fundaci´on de la UAM” accessi-

ble technologies project titled: “AssisT-Out: a mo-

bile guiding tool for outdoors movements adapted to

people with cognitive disabilities” and “e-Training

y e-Coaching para la integraci´on socio–laboral”

(TIN2013–44586–R)

HEALTHINF2015-InternationalConferenceonHealthInformatics

350

REFERENCES

105th Congress of the United States (1998). Assistive tech-

nology act of 1998. Pub.L. 105?394, 112 Stat. 3627,

S. 2432, enacted November 13, 1998.

Barbieri, T., Bianchi, A., Fraternali, P., and Tacchella, C.

(2010). Autonomamente project-design, implemen-

tation and evaluation of a multimodal domotic appli-

cation to support persons with cognitive disabilities.

In e-Health Networking Applications and Services

(Healthcom), 2010 12th IEEE International Confer-

ence on, pages 1–6. IEEE.

Beeharee, A. K. and Steed, A. (2006). A natural wayﬁnd-

ing exploiting photos in pedestrian navigation sys-

tems. In Proceedings of the 8th conference on Human-

computer interaction with mobile devices and ser-

vices, pages 81–88. ACM.

Boisvert, A., Paquette, L., Pigot, H., and Giroux, S. (2009).

Design challenges for mobile assistive technologies

applied to people with cognitive impairments. Ambi-

ent Assistive Health and Wellness Management in the

Heart of the City, pages 17–24.

Braddock, D., Rizzolo, M., Thompson, M., and Bell, R.

(2004). Emerging technologies and cognitive dis-

ability. Journal of Special Education Technology,

19(4):49–56.

Brown, C. M. (1999). Human-computer interface design

guidelines. Intellect Books.

Caballero-Hern´andez, R., Moreno, J., Molina, A., Celma,

S., S´anchez, J., Carri´on, R., Casado, E., Rodr´ıguez,

R., Pulido, A., Tafalla, C., et al. (2012). 2d-tasks for

cognitive rehabilitation. In 5th European Conference

of the International Federation for Medical and Bio-

logical Engineering, pages 838–841. Springer.

Carmien, S. (2003). Maps: Dynamic scaffolding for in-

dependence for persons with cognitive impairments.

User Modeling 2003, pages 147–147.

Carmien, S. (2006). Assistive technologies for persons with

cognitive disabilities - artifacts of distributed cogni-

tion. In CHI 2006 workshop: Designing Technology

for People with Cognitive Impairments.

Carmien, S. (2010). Socio-technical environments and as-

sistive technology abandonment.

Cole, E. (2011). Patient-centered design: interface person-

alization for individuals with brain injury. Universal

Access in Human-Computer Interaction. Users Diver-

sity, pages 291–300.

Fickas, S., Sohlberg, M., and Hung, P.-F. (2008). Route-

following assistance for travelers with cognitive im-

pairments: A comparison of four prompt modes.

International Journal of Human-Computer Studies,

66(12):876–888.

Garc´ıa de Marina, A. G., Carro, R. M., and Haya, P. (2012).

Where should i go?: guiding users with cognitive lim-

itations through mobile devices outdoors. In Proceed-

ings of the 13th International Conference on Inter-

acci´on Persona-Ordenador, page 46. ACM.

Garc´ıa-Herranz, M., Olivera, F., Haya, P., and Alam´an, X.

(2012). Harnessing the interaction continuum for sub-

tle assisted living. Sensors, 12(7):9829–9846.

Hegarty, M., Richardson, A. E., Montello, D. R., Lovelace,

K., and Subbiah, I. (2002). Development of a self-

report measure of environmental spatial ability. Intel-

ligence, 30(5):425–447.

Hidalgo, E., Castillo, L., Madrid, R., Garc´ıa-P´erez,

O.,

Cabello, M., and Fdez-Olivares, J. (2011). Athena:

Smart process management for daily activity planning

for cognitive impairment. Ambient Assisted Living,

pages 65–72.

Ishikawa, T., Fujiwara, H., Imai, O., and Okabe, A. (2008).

Wayﬁnding with a gps-based mobile navigation sys-

tem: A comparison with maps and direct experience.

Journal of Environmental Psychology, 28(1):74–82.

Kr¨uger, A., Aslan, I., and Zimmer, H. (2004). The effects of

mobile pedestrian navigation systems on the concur-

rent acquisition of route and survey knowledge. Mo-

bile Human-Computer Interaction–MobileHCI 2004,

pages 39–60.

Lemoncello, R., Moore Sohlberg, M., and Fickas, S. (2010).

How best to orient travellers with acquired brain in-

jury: A comparison of three directional prompts.

Brain Injury, 24(3):541–549.

Lepist¨o, A. and Ovaska, S. (2004). Usability evalua-

tion involving participants with cognitive disabilities.

In Proceedings of the third Nordic conference on

Human-computer interaction, pages 305–308. ACM.

Lewis, C. (1982). Using the” thinking-aloud” method in

cognitive interface design. IBM TJ Watson Research

Center.

Liu, A., Hile, H., Borriello, G., Brown, P., Harniss, M.,

Kautz, H., and Johnson, K. (2009). Customizing di-

rections in an automated wayﬁnding system for indi-

viduals with cognitive impairment. In Proceedings of

the 11th international ACM SIGACCESS conference

on Computers and accessibility, pages 27–34. ACM.

Lund, A. M. (2001). Measuring usability with the use ques-

tionnaire. Usability interface, 8(2):3–6.

Mihailidis, A., Blunsden, S., Boger, J., Richards, B., Zutis,

K., Young, L., and Hoey, J. (2010). Towards the devel-

opment of a technology for art therapy and dementia:

Deﬁnition of needs and design constraints. The Arts

in Psychotherapy, 37(4):293–300.

Mihailidis, A., Fernie, G., and Cleghorn, W. (2000). The

development of a computerized cueing device to help

people with dementia to be more independent. Tech-

nology and Disability, 13(1):23–40.

O’Neill, B. and Gillespie, A. (2008). Simulating naturalis-

tic instruction: the case for a voice mediated interface

for assistive technology for cognition. Journal of As-

sistive Technologies, 2(2):22–31.

Richter, K.-F., Dara-Abrams, D., and Raubal, M. (2010).

Navigating and learning with location based services:

A user-centric design. In Proc. 7th International Sym-

posium on LBS & TeleCartography, pages 261–276.

Richter, K.-F. and Duckham, M. (2008). Simplest instruc-

tions: Finding easy-to-describe routes for navigation.

Geographic Information Science, pages 274–289.

Siegel, A. W. and White, S. H. (1975). The development

of spatial representations of large-scale environments.

Advances in child development and behavior, 10:9.

DesignConsiderationsandEvaluationMethodologyforAdaptedNavigationalAssistantsforPeoplewithCognitive

Disabilities

351