Assisting Speech Therapy for Autism Spectrum Disorders with an

Augmented Reality Application

Camilla Almeida da Silva

, Ant´onio Ramires Fernandes

and Ana Paula Grohmann

Faculdade de Computac¸˜ao e Inform´atica, Universidade Presbiteriana Mackenzie, S˜ao Paulo, Brazil

Department of Informatics, University of Minho, Braga, Portugal

Associac¸˜ao para a Inclus˜ao e Apoio ao Autista, Braga, Portugal

Keywords:

Autism Spectrum Disorder, Augmented Reality, Augmented and Alternative Communication, Human-

computer Interaction.

Abstract:

Graphics based systems of Augmented and Alternative Communication are widely used to promote commu-

nication of people with Autism Spectrum Disorders. However, studies indicate the inability of some of these

people in understanding the used symbols. This study discusses an integration of Augmented Reality in com-

munication interventions, by relating elements of Augmented and Alternative Communication and Applied

Behaviour Analysis strategies. An Augmented Reality based interactive system to support interventions is

discussed, and the report of its usage in interventions with children suffering from Autism Spectrum Disorders

is presented.

1 INTRODUCTION

Children with Autism Spectrum Disorders (ASD) are

affected with various impairments in communication,

social interaction and imagination, three major com-

ponents of self-development. Regarding communi-

cation issues, these range from the total absence of

language, 20% to 30% (Klin, 2006), to the lack of ef-

fectiveness in the communication process, reinforcing

the need to promote initiatives to improve communi-

cation skills.

Applied Behaviour Analysis (ABA) interventions

for autism in early childhood are an effective prac-

tice to improve socially relevant behaviours and teach

new skills, through several established teaching tools

and positive reinforcement strategies. Augmented

and Alternative Communication (AAC) based ap-

proaches, in particular those based on graphics, are

the most used in interventions in children with ASD,

being highly relevant in communication promotion,

reduction of behavioural problems, and environment

awareness (Committee on Educational Interventions

for Children with Autism, 2001).

The graphic based system usage is supported by

the strong visual processing ability seen in many chil-

dren with ASD. Nevertheless, imagination impair-

ments caused by rigidity and lack of imagination can

prevent the usage of some symbols (Herrera et al.,

2012). In (Herrera et al., 2012) it is argued that some

children with ASD, when confronted with a symbol,

only capture a set of lines, shapes and colours, and the

usage of communication symbols would lead to mem-

orization and context association. The authors show

that changes in the background or width of the lines

are sufﬁcient to prevent recognition on a previously

learned symbol. Hence, they conclude that these sym-

bols are meaningless to those children. In (Commit-

tee on Educational Interventions for Children with

Autism, 2001) it is stated that the lack of symbolic

capacity is one of the main handicaps in communica-

tion, reﬂecting the difﬁculty in learning conventional

or shared meanings of symbols. To achieve commu-

nication, symbols must be meaningful to the child

(Avila, 2011).

Computer animations, 3D graphics and sounds,

may provide deeper engagement for children in their

activities as suggested in several studies on the bene-

ﬁts of Information Technologies in interventions with

children with ASD (Moore et al., 2000; Moore and

Calvert, 2000; Goldsmith and LeBlanc, 2004). The

use of tangible interfaces has also been suggested by

(Rafﬂe et al., 2009) to promote alternative ways of

collaboration and communication, engaging the chil-

dren for longer periods of time than with regular in-

terventions.

Augmented Reality (AR), merging virtual ob-

Almeida da Silva C., Ramires Fernandes A. and Grohmann A..

Assisting Speech Therapy for Autism Spectrum Disorders with an Augmented Reality Application.

DOI: 10.5220/0004893200380045

In Proceedings of the 16th International Conference on Enterprise Information Systems (ICEIS-2014), pages 38-45

ISBN: 978-989-758-029-1

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

jects into real environments, can be explored in

some intervention areas for ASD children, such as

self-awareness, augmented communication, emotion

awareness and identiﬁcation, social commitment and

concept development (Herrera et al., 2006). These vi-

sually oriented approaches are highly suitable due to

the remarkable ability of most autistic people to excel

at visual spatial skills (Grandin, 2009).

Considering the relevance of AAC approaches,

and the impact of symbols to promote communica-

tion, we performed a study on the usage of AR to

assist interventions and expand language with chil-

dren with ASD. The interactive AR system devel-

oped in this study uses both AAC and ABA prin-

ciples to support interventions focused on facilitat-

ing graphic symbol comprehension. The system re-

inforces the communication symbols superimposing

virtual objects and animations over the real environ-

ment screen view.

The system’s human-computer interaction was

previously discussed with speech-language therapists

to ensure its suitability for both therapists and chil-

dren. The system’s features include the creation of

activities with AR graphics and sound to enrich the

symbol’s meaning, with both visual and audio rein-

forcements. The system is adaptable to each individ-

ual child, allowing to have multiple activity conﬁg-

urations, and supports a number of different activity

templates to increase the options for the therapist.

In order to evaluate the system, the application

was tested in speech-language sessions with four chil-

dren with ASD. The tests fulﬁlled the framework pro-

posed by Moore (Moore et al., 2000), where computer

assisted learning systems for ASD children should fo-

cus on at least one on the major impairments; the

projects should be based on current established prac-

tices; and evaluation should be performed in coopera-

tion with actual therapists.

2 AUTISM SPECTRUM

DISORDER

ASD is a group of developmental brain disorders af-

fecting individuals from all races and cultures. It

presents a wide range of possible symptoms and vary-

ing degrees of severity, the common feature being an

early developmental disruption in socialization pro-

cesses (Klin, 2006).

The Centers for Disease Control and Prevention

(Centers for Disease Control and Prevention, 2013)

estimate that about 1% of individuals have ASD in

Europe, Asia, and North America. ASD is more com-

mon among boys than among girls, with a ratio of 5

to 1.

2.1 Interventions

The main goal of an intervention is to promote func-

tional and spontaneous communication, social in-

teraction and symbolic play to increase functional

independence and quality of life. These interven-

tions address communication, social skills, unsuit-

able behaviours, among other common issues caused

by ASD (Myers et al., 2007). The intervention

programs require a multidisciplinary base, involving

behavioural therapy, educational programs, speech-

language and communication therapy (Gadia et al.,

2004).

Applied Behavioural Analysis methods, such as

intense behavioural interventions, directed towards

long term behaviour modiﬁcation, have as goals to

increase and maintain adequate behaviours, reduce

maladaptive behaviours, and teach new skills (Myers

et al., 2007). These methods have a short duration

and are effective in both adults and children (Com-

mittee on Educational Interventions for Children with

Autism, 2001). Discrete Trial Training (DTT) is also

an ABA method to enhance skills such as attention,

compliance, imitation, and discrimination learning,

among others (Smith, 2001). It consists of a struc-

tured learning method whereby a complex learning

sequence is divided in small steps, with reinforce-

ments and assistance by the therapist where required.

The American Speech-Language-HearingAssoci-

ation deﬁnes AAC as an area of clinical practice fo-

cused on impairment and disability patterns of indi-

viduals with severe expressive communication disor-

ders (Sevcik and Romski, 2000). AAC methods and

tools are also highly adaptable to the autistic child

needs and personality.

Graphics based systems take advantage of the

strong visual processing skills found in many in-

dividuals with ASD, and have been proven effec-

tive, through functional communication training, in

increasing communication reception in small chil-

dren, and to replace disruptive behaviour such as ag-

gression, self-aggression and crying (Committee on

Educational Interventions for Children with Autism,

2001). The environment, space, and time concepts

are other examples where AAC can be used.

The Picture Exchange Communication System

(PECS), developed by Bondy and Frost in 1995

(Bondy and Frost, 1994), is an AAC method for

functional communication based on picture exchange.

PECS follows ABA principles being structured in a

sequence of steps. Initially the child is taught to initi-

ate requests through ﬁgures. Later the child will start

AssistingSpeechTherapyforAutismSpectrumDisorderswithanAugmentedRealityApplication

building sentences, answering questions and, on latter

stages, to make comments. Several studies (Charlop-

Christy et al., 2002; Kravits et al., 2002; Ganz and

Simpson, 2004) have shown the beneﬁts of PECS, re-

sulting in vocabulary acquisition and social behaviour

improvements.

3 RELATED WORK

VR and AR techniques have been used extensively to

assist interventions for children with ASD.

Project INMER-II (Belen, 2004) was designed

based on an a serious game in a VR environment

related to shopping activities. The results show in-

creased levels of functional and symbolic comprehen-

sion, and imagination. Furthermore, one child was

able to successfully generalize the experience to a real

environment.

Virtual avatars with emotional facial expressions

were used in project Virtual Messenger (Fabri and

Moore, 2005). The goal was to help users to rec-

ognize facial expressions. In the ﬁrst stage, the user

would be asked to select the emotion corresponding

to the avatar presented on screen, or to select a facial

expression that corresponds to an emotion. The sec-

ond level requires the user to select an emotion that is

appropriate for a scenario presented on screen. When

in the third level, the user is presented with an avatar’s

facial expression, and is required to select a scenario

that could cause the avatar’s expression. The results

show that people with ASD were able to increase their

level of understanding of emotions based on facial ex-

pressions.

Project AS Interactive was developed to support

people with ASD to improve social skills through

VR environments simulating a cafeteria and a bus

(Parsons et al., 2006). The study performed with

teenagers with ASD led to the conclusion that these

simulated environments could be successfully used to

help people with ASD to engage in social interactions.

In (Herrera et al., 2006) a system based on AR

with infrared markers was described. The system

would superimpose the card symbols on the user’s

image in real time. However, this prototype was

never tested on children with ASD. Later, in (Her-

rera et al., 2012) a AR system using Microsoft Kinect

was proposed, having as main goals teaching self-

consciousness, body scheme and postures, communi-

cation and imitation, using serious games to promote

interaction with the children.

Tentori and Hayes proposed the Mobile Social

Compass framework for the development of mobile

AR systems. The Social Compass Curriculum is a be-

havioural and educational curriculum for social skills

training, based on stories and visual paper clues to

guide the child in both active and passive social inter-

actions. The framework focus in the concept of Inter-

action Immediacy, providing a set of visual clues to

assist the child in anticipating situations (Tentori and

Hayes, 2010).

MOSOCO is a mobile application implemented

using the above described framework, using AR and

visual guidance for children with ASD, to assist chil-

dren with ASD in social interactions in real life. The

system guides the children in the basic social skills

deﬁned in The Social Compass curriculum, encour-

aging them through interactive resources. The sys-

tem’s evaluation was performed with three children

and the results show that the system increased so-

cial interactions, both in quantity and quality, mitigat-

ing behavioural and social disruptions, and allowed

the social integration of these children in neurotypi-

cal groups.

Another mobile AR application, Blue’s Clues,

was proposed in (Escobedo and Tentori, 2011), to

help children with ASD moving from one place to

another. Blue’s Clues concept was developed for a

school environment, aiming at providing the neces-

sary mobility instructions, visual and audio clues, to

guide students with ASD.

4 THE AR BASED PROTOTYPE

In this section we propose an AR based system that

allows the creation of interactive activities supported

on AAC and ABA principles to assist interventions

with children with ASD, aided by a therapist.

The goal was not to impose a new methodology to

the therapist, but to offer a support tool that could en-

rich the current daily activities in the speech-language

sessions. Hence, all the developmentwas initially dis-

cussed with speech-language therapists, to ensure that

the prototype would ﬁt painlessly in the current activ-

ities, and that its interface was clear and intuitive for

both the therapist and the children.

The prototype allows the design of activities for

each individual child attending to the difference in de-

gree of the ASD, the individual personality, and the

previous activities developed by the therapist. Within

each activity, we aimed at providing as much ﬂexibil-

ity as possible. Several templates were provided for

this initial evaluation to guide the therapist in design-

ing the individual activities.

The usage of AR is directed towards superimpos-

ing 3D models and animations on cards with commu-

nication symbols that the child holds or places on the

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

table. This creates a richer experience, and takes ad-

vantage of the visual skills found in many children

with ASD.

The selection of AR as the underlying technology

for this project aims at:

1. Promote a natural interaction with the computer

through the use of tangible interfaces;

2. Facilitate the generalization of the acquired

knowledge;

3. Allow the enrichment of traditional interventions

based on communication symbols without impos-

ing a new methodology on the therapy;

4. Eliminate the distractions caused by traditional in-

terfaces such as mouse and keyboard.

In (Billinghurst et al., 2005) is suggested that for

the use of tangible interfaces to be intuitive and natu-

ral, it is recommended that the selection of physical

objects and interaction metaphors should be famil-

iar to users, allowing them to interact with the sys-

tem relying on previous skills and experiences. In our

project we use communication symbols which are al-

ready familiar to the children. The child interfaces

with the system by showing cards with communica-

tion symbols, which is a natural behaviour for the

child under intervention. Furthermore, the acquired

knowledge is implicitly associated with the commu-

nication symbols, hence, the child should be able to

recall its experiences afterwards.

The prototype was built with open software tools

and APIS such as irrKlang for audio, ARToolkit for

the AR component and MySQL. These options aim

at providing a low cost solution, and multiplatform

future support.

4.1 The Back Ofﬁce

The activity construction process takes place in a sim-

ple back ofﬁce support application where the therapist

can select a template and design interactive activities

to explore with the children. Figure 1 shows a screen-

shot of the activity construction stage.

Initially the therapist associates words with sym-

bol cards, 3D graphics, sounds and animations, and

builds phrases with word sequences. The words can

be further deﬁned as belonging to user-deﬁned cate-

gories such as actions, fruits, or animals.

The data set of words, and their associated me-

dia, is then used to construct the activities based on

the available templates. Each template provides full

parameterization including reinforcement deﬁnition,

and error management.

4.2 Activity Templates

The available templates are: Free, Category, Discrim-

inate, and Phrase. These are based on stages I (Free),

III (Category and Discriminate), and IV (Phrase) of

PECS. The possibility to use different templates also

seeks to avoid having the autistic child engaging in

repetitive tasks for long periods of time.

The basic procedure consists in showing a card in

the ﬁeld of view of the web cam. The recognition of

the pattern in the card will trigger an action.

The Free template, the simplest one, allows both,

child and therapist, to interact freely with the proto-

type. When designing an activity with this template

the therapist can associate the action to be triggered

when the card is shown or occluded. This action,

which can be a graphic, an animation, or a sound, can

act as a positive reinforcement depending on the in-

tervention context, or just be used to engage the child

into exploring the system. Additional actions can be

speciﬁed when the card is occluded. As an exam-

ple assume an activity based on cards with animals.

When the child shows a card the system may display

on screen a 3D model of the animal on top of the card.

If the child covers the card with the hand then the an-

imal sound can be played.

In the Discriminate template the child is required

to identify an object based on a clue provided on

screen, and present the card that matches the clue.

Positive and negative reinforcements can be associ-

ated with each individual clue. The system supports

the conﬁguration for special cards that the therapist

can include in the activity. An ’hint’ card will trig-

ger a user-deﬁned hint when shown to the system. A

’next’ card tells the system to move on to the next

object. Finally the ’end’ card will end the activity.

These cards, if provided by the therapist, allow for a

freer interaction from the child, since the child can

control the activity when possessing all three special

cards, or some subset. The special cards can also be

used by the therapist to control the activity.

The Category template can be conﬁgured in two

different ways. Initially the system will present a hint

on screen, namely a graphic or sound. The child is

then requested to select and show a card either with

the respective category or with an item in the same

category. Positive reinforcements, applied when the

answer is correct, are conﬁgurable, as well as the ac-

tions to be triggered for showing the wrong card.

In the Phrase template the child is requested

to show a sequence of cards, which repre-

sent the subject, verb and noun in a phrase,

see Figure 2 for an example using symbols

from the ARASAAC collection (available at

AssistingSpeechTherapyforAutismSpectrumDisorderswithanAugmentedRealityApplication

Figure 1: Screen shot of back ofﬁce interface for the template conﬁguration.

http://www.catedu.es/arasaac/descargas.php). When

each card is placed on the table a reinforcement can

be triggered by the system. This reinforcement will

hint the child if the right card is being used, and also

if the card is in the correct position within the phrase.

The system will detect when a phrase is complete and

trigger the user-deﬁned positive reinforcement.

Figure 2: An example of a phrase.

5 TESTING THE PROTOPYPE

The test was performed in an association that sup-

ports children with ASD, the Associac¸˜ao para a In-

clus˜ao e Apoio ao Autista, located in Braga, Portugal

(http://www.aia.org.pt/).

The prototype was tested by a speech-language

therapist, the third author, that performs interventions

regularly with children, and four children diagnosed

with ASD. The goal was to test if the software was

clear to the therapist, and to evaluate its usage as a

supporting tool in the interventions performed.

The therapist had 13 sessions with four boys diag-

nosed with ASD. The children were between the ages

of 6 and 10, had some level of orality, and attended

school.

Some children were previously assessed as par-

ticipative in the speech-language sessions, while oth-

ers have an extremely passive behaviour and barely

engaged in spontaneous interaction and communica-

tion with the therapist. One child had several verbal

and motor stereotypies and expressed himself mainly

through echolalia.

A qualitative study was performed with data gath-

ered through observation by the team, including the

therapist, in order to evaluate the suitability of the

software usage by the children, namely what be-

haviours were triggered by the usage of AR, and if the

children had beneﬁted from the interaction with the

software. To support the study we recorded the ses-

sions in video, and ﬁlled observationgrids with the in-

teraction stages as well as observations that emerged

in the sessions with the children (Figure 3).

Figure 3: Children interacting with the prototype.

5.1 Strategies

Based on the previous history with the children, as

well as the existing intervention plan for each child,

the therapist set the following scripts for these ses-

sions.

For three children the goal was set to: (1) iden-

tify animals, nutrients, and daily objects; (2) Identify

and designate gender; (3) Identify and designate ac-

tions; (4) Discriminate and identify sounds; (5) An-

swer yes/no questions, together with the respective

head nod or shake; and (6) Build sentences with sub-

ject, verb and noun.

The other child had as goals to: (1) Enhance

declarative skills; (2)Improve language skills; and (3)

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

Create and imagine using language.

The therapist created six activities based on the

four available templates to achieve these goals.

5.2 Evaluation

In general the children were highly motivated by the

usage of AR. Superimposed 3D models and computer

animations on top of cards displaying familiar sym-

bols caused a higher degree of engagement in the ac-

tivities, and had the children requesting the therapist

to ’play’ with the software when arriving.

The child with stereotypies, when in traditional in-

terventions, used to keep asking to go to the window

to watch for cars passing by, and kept repeating out

of context phrases. After the AR based session had

ended, when they were ready to start a different ac-

tivity (not related to this project), the child asked the

therapist for the dog marker and autonomously pre-

sented it to the system. A joyfully expression emerged

when the dog’s bark was heard.

A child deﬁned by the therapist as passive, re-

garding communication and behaviour, revealed to be

highly motivated, repeating everything the computer

’said’.

The activities were easily grasped by the children,

possibly due to the use of tangible interfaces based on

the cards with symbols, something which was highly

familiar to these children. The prototype did not re-

quire any new form of communication, showing cards

was something that was already present in the ses-

sions prior to this test. Hence, the interaction with the

computer presented itself as familiar to the child.

After 13 sessions the therapist reported her im-

pressions. On a positive note, the therapist stated that

the children kept the interest on the activities for a

longer period of time, and inclusively that children

had the initiative to ask and use the software on their

own. The children acquired new vocabulary and con-

sistently managed to perform correctly the proposed

activities.

However, the therapist also noted that a particu-

lar child would sometimes disregard everything else

concentrating only on his image on the screen. The

joint attention was not as high and eye contact oc-

curred less often, suggesting that a balance must be

achieved between the several different types of inter-

ventions and activities.

The therapist concluded that an AR based ap-

proach has the potential to present beneﬁts in speech-

language therapy although the social area may be ne-

glected with some children.

5.3 Detected Issues

Some issued came up during the test which are worth

mentioning in this report. First we had some situa-

tions where a card would be incorrectly recognized,

triggering an inappropriate action. Other times an ac-

tion was triggered without having any card on display.

These issues came up due to the variations in lighting

during the day.

One child actually enjoyed the fact that the sys-

tem made ’mistakes’ and was eager to correct these

mistakes. The therapist also took these system mis-

classiﬁcations and explored them within the interven-

tion leading to an even more engaging session. These

unintended recognition failures must be ﬁxed, as chil-

dren with more rigid cognitive processes may be un-

able to deal with these situations. Nevertheless, this

obviously leads to the question: should the system fail

occasionally on purpose? Should this feature be intro-

duced in the conﬁguration?

On another session when the system triggered a

sound without a card being shown, the child imme-

diately sought for the associated card and showed it

to the system. Again, while these unattended failures

must be avoided, they may become part of an activity

with some children, introducing an element of sur-

prise and breaking the pattern of the activity.

The team also noted that the Phrase template

needed to be redesigned to include variations and

more reinforcements. The system should be able to

provide hints regarding the type of the missing ele-

ments and their relative position. Furthermore, the

system should allow reinforcements when, although

all the correct cards are in the table, the order is not

correct. Another option to explore is to start with in-

complete phrases and ask the child to identify the type

of the missing element, as well as the speciﬁc missing

card.

The last issue is related to the child’s interaction

when motor stereotypies are present. In our study a

child needed assistance from the therapist to be able

to show the cards. This issue requires support for a

different form of interaction.

6 CONCLUSIONS

The present work has proposed an approach, relating

elements from AAC and ABA, to apply AR in com-

munication and language interventions with children

with ASD to mitigate the issues caused by the impair-

ments in communication skills and imagination.

The resulting prototype has been qualitatively

evaluated in an association that supports children with

AssistingSpeechTherapyforAutismSpectrumDisorderswithanAugmentedRealityApplication

ASD and has therapists conducting interventions in

many areas. The sessions were conducted by a ther-

apist and the behaviours of the children and their in-

teraction with the system was recorded on video and

using observation grids.

A common issue in qualitatively evaluations re-

lates to the generalization of results to the remain-

ing population. We have no intention of suggesting

that the data presented in here can be generalized.

This is particularly relevant in the case of ASD due

to the wide range of disorders and their varying de-

grees. This study presents only a contribution to the

development of technological systems to support in-

terventions.

Our study shows that the usage of information

technologies, in particular AR based systems, in

speech-language interventions, complementing and

supporting the traditional approaches is an option that

must be further explored. The results suggest that the

system usage could represent an added value since we

could observe greater motivation, word acquisitions,

and commitment in performing correctly, as observed

by the therapist.

Due to the use of AR and tangible interfaces,

familiar physical objects, and an interaction proce-

dure similar to the one in the previous AAC sessions,

where cards are exchanged between child and thera-

pist, the system was easily grasped by the children.

As expected, the children were able to use their past

experience to interact effortlessly with the system and

hopefully they will be able to generalize the acquired

knowledge.

We believe that the positive results obtained in this

study are, to some extent, due to the engaging envi-

ronment provided by the application, and its tangible

interface. These factors stimulate cognitive process-

ing and lead to visual learning.

These results illustrate the potential of AR sys-

tems to be used in linguistic skill development due

to the signiﬁcant increase in interaction and commu-

nication initiated by the children.

While technologies such as the ones in our proto-

type can provide very positive results, we must pay

particular attention to avoid having the child com-

pletely immersed on some application details and ne-

glecting social interaction. The therapist plays a fun-

damental role in these augmented ’interventions’.

6.1 Future Work

We believe that the results presented show the need

for deeper and further studies, adding more interac-

tion templates, extending their conﬁguration, and test-

ing the system with more children.

Game oriented activity templates could provide an

extra level of engagement for the children. Adding

templates should be explored to increase the range

of possibilities for the therapist to conduct the inter-

ventions. The phrase template, as mentioned before,

would also beneﬁt from a subdivision or at least more

conﬁguration options.

The card recognition subsystem also needs to be

ﬁne-tuned to increase robustness, hence suppressing

unintentional card misclassiﬁcation. These system er-

rors, due to the way children reacted to them, raise an

interesting question that we believe is worth pursuing:

Would a system that intentional provides a wrong an-

swer be a way of further engaging at least some chil-

dren? What is a ’good’ wrong answer?

Children with motor stereotypies also require a

different interaction procedure. This condition makes

it harder for the system to correctly classify the cards.

This is an important issue since a signiﬁcant number

of children with ASD suffer from motor stereotypies.

In this study we focused on speech-language in-

terventions. We would like to explore other types of

interventions, such as behavioural and sensory thera-

pies.

Finally, the limited number of participating chil-

dren does not reﬂect the broad range of the spectrum,

nor do they allow for any generalization in a particu-

lar subset of these disorders. While the generalization

for the whole spectrum of disorders is an impossible

task, more participating children would allow for a

more relevant analysis, regarding the signiﬁcance of

the contribution of a system such as the one presented

in here when used in interventions.

ACKNOWLEDGEMENTS

Work partially funded by National Funds through

FCT - Fundac¸˜ao para a Ciˆencia e a Tecnologia

(Portuguese Foundation for Science and Technol-

ogy) within project PEst-OE/EEI/UI0752/2011 and

by Conselho Nacional de Desenvolvimento Cient´ıﬁco

e Tecnol´ogico (CNPq-Brazilian National Council for

Scientiﬁc and Technological Development) within

program Ciˆencia sem Fronteiras.

REFERENCES

Avila, B. G. (2011). Comunicac¸˜ao Aumentativa e Alterna-

tiva para o Desenvolvimento da Oralidade de Pessoas

com Autismo. Master’s thesis, Universidade Federal

do Rio Grande do Sul, Porto Alegre, Brasil.

ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems

Belen, S. O. (2004). INMER-II: Sistema de Inmersi´on en

Realidad Virtual para Personas con Autismo. Tec-

noneet 2004: Retos y realidades de la Inclusi´on Digi-

tal.

Billinghurst, M., Grasset, R., and Looser, J. (2005). Design-

ing augmented reality interfaces. SIGGRAPH Com-

put. Graph., 39(1):17–22.

Bondy, A. S. and Frost, L. A. (1994). The picture exchange

communication system. Focus on Autism and Other

Developmental Disabilities, 9(3):1–19.

Centers for Disease Control and Prevention (2013).

Autism spectrum disorders. http://www.cdc.gov/

ncbddd/autism/data.html. Accessed: 2013-10-15.

Charlop-Christy, M. H., Carpenter, M., Le, L., LeBlanc,

L. A., and Kellet, K. (2002). Using the picture ex-

change communication system (pecs) with children

with autism: Assessment of pecs acquisition, speech,

social-communicative behavior, and problem behav-

ior. Journal of applied behavior analysis, 35(3):213–

231.

Committee on Educational Interventions for Children with

Autism, N. R. C. (2001). Educating Children with

Autism. The National Academies Press.

Escobedo, L. and Tentori, M. (2011). Blue’s clues: An aug-

mented reality positioning system. Child Computer

Interaction Workshop (CHI).

Fabri, M. and Moore, D. (2005). The use of emotionally ex-

pressive avatars in collaborative virtual environments.

Virtual Social Agents, page 88.

Gadia, C. A., Tuchman, R., and Rotta, N. T. (2004). Autism

and pervasive developmental disorders. Jornal de Pe-

diatria, 80:83 – 94.

Ganz, J. B. and Simpson, R. L. (2004). Effects on com-

municative requesting and speech development of the

picture exchange communication system in children

with characteristics of autism. Journal of Autism and

Developmental Disorders, 34(4):395–409.

Goldsmith, T. R. and LeBlanc, L. A. (2004). Use of

technology in interventions for children with autism.

Journal of Early and Intensive Behavior Intervention,

1(2):166–178.

Grandin, T. (2009). Thinking in Pictures. Bloomsbury Pub-

lishing.

Herrera, G., Casas, X., Sevilla, J., Luis, R., Pardo, C., and

Plaza, J. (2012). Pictogram room: Natural interac-

tion technologies to aid in the development of children

with autism. Annuary of Clinical and Health Psychol-

ogy, 8:39–44.

Herrera, G., Jordan, R., and Gimeno, J. (2006). Explor-

ing the advantages of augmented reality for interven-

tion in asd. In Autism Safari 2006 - 2nd Word Autism

Congress.

Klin, A. (2006). Autism and asperger syndrome: an

overview. Rev Bras Psiquiatr, 28(Supl I):S3–11.

Kravits, T. R., Kamps, D. M., Kemmerer, K., and Po-

tucek, J. (2002). Brief report: Increasing commu-

nication skills for an elementary-aged student with

autism using the picture exchange communication

system. Journal of autism and developmental disor-

ders, 32(3):225–230.

Moore, D., McGrath, P., and Thorpe, J. (2000). Computer-

aided learning for people with autism; a framework

for research and development. Innovations in Educa-

tion and Training International, 37(3):218–228.

Moore, M. and Calvert, S. (2000). Brief report: Vocabu-

lary acquisition for children with autism: Teacher or

computer instruction. Journal of Autism and Develop-

mental Disorders, 30(4):359–362.

Myers, S. M., Johnson, C. P., and the Council on Chil-

dren With Disabilities (2007). Management of chil-

dren with autism spectrum disorders. Pediatrics,

120(5):1162–1182.

Parsons, S., Leonard, A., and Mitchell, P. (2006). Vir-

tual environments for social skills training: comments

from two adolescents with autistic spectrum disorder.

Computers and Education, 47(2):186 – 206.

Rafﬂe, H., Farr, W., and Yuill, N. (2009). Collaborative ben-

eﬁts of a tangible interface for autistic children. ACM

SIGCHI.

Sevcik, R. and Romski, M. (2000). A.A.C.: More

Than Three Decades of Growth and Develop-

ment. http://www.asha.org/public/speech/disorders/

AACThreeDecades/. Accessed: 2013-10-15.

Smith, T. (2001). Discrete trial training in the treatment

of autism. Focus on autism and other developmental

disabilities, 16(2):86–92.

Tentori, M. and Hayes, G. R. (2010). Designing for inter-

action immediacy to enhance social skills of children

with autism. In Proceedings of the 12th ACM inter-

national conference on Ubiquitous computing, pages

51–60. ACM.

AssistingSpeechTherapyforAutismSpectrumDisorderswithanAugmentedRealityApplication