Educational Robotics as a Learning Aid for Disabled Children

Educational Robotics as a Learning Aid for Disabled Children

Carlos Valadão1, Teodiano Freire Bastos2, Magdo Bôrtole3, Victor Perim4, Daniel Celino5, Fauzi Rodor6, Agda Gonçalves7, Humberto Ferasoli8

[email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]

Electrical Engineering Post-Graduation Program Universidade Federal do Espírito Santo1,2,3,4,5,6,7

Fernando Ferrari Av., 514 29075-910, Vitória, Brazil

Universidade Estadual Paulista8

Av. Eng. Luiz Edmundo C. Coube 14-01

17033-360 Bauru – SP, Brazil

Abstract - Severe disabled children have little chance of environmental and social exploration and discovery, and due this lack of interaction and independency, it may lead to an idea that they are unable to do anything by themselves. This idea is called “learned helplessness” and is very negative for the child cognitive development and social development as well. With this entire situation it is very likely that the self-steam and mood of this child.

Trying to help these children on this situation, educational robotics can offer and aid, once it can give them a certain degree of independency in exploration of environment. The system developed in this work allows the child to transmit the commands to a robot. Sensors placed on the child’s body can obtain information from head movement or muscle pulses to command the robot to carry the tasks. Also, this system can be used with a variety of robots, being necessary just a previous configuration. It is expected that, with the usage of this system, the disabled children have a better cognitive development and social interaction, balancing in a certain way, the negative effects of their disabilities.

Keywords - disabled children, educational robotics, learned helplessness, cognitive development, learning aid.

I. INTRODUCTION

Learning in childhood is done by exploration and discovery of the environment where the child lives. According to Piaget’s definition, up to the second year of life the child is living the sensorial motor period [4,1]. During this stage of development, interaction with the environment is done through physical sensation and body stimulus [4,7]. This idea means that they learn how to interact with their own body as well as the environment by repeating experiences and exploring the world through their senses. At the end of this sensorial-motor period the children has the notion of space, position of objects inside the space and time, and some relation among them [4].

This spatial object manipulation and environmental interaction are fundamental for the child cognitive development and the ones with severe disabilities are blocked from experiencing the world as the other children do. With this situation they may become delayed in terms of learning through exploration by themselves.

Considering this lack of independency, exploration and spontaneity on discovery of the environmental area around the child, those ones can have a negative influence towards learning and social interaction. All this situation of lack of stimulus can provoke a late childhood [4].

Due to these body-limitations, disabled children are, usually, very dependent on their parents or caretakers to interact with the world. Brinker and Lewis [4,5] suggest

that the child’s behavior can, by itself, dexperiences their parents and caretakers woto have. These choices may restrict ecognitive and social development of the infa

This entire situation can create a lackexploration and consequently also develolearned helplessness, which they see themsto do anything independently or withoutWith this idea the child usually adopts a paof interest behavior towards the world he or

All these elements can compromise the bbefore, once they become socially passives To minimize this, it is necessary that the chexplore the world, through alternative according to what the child has to develothis is very likely that this child can motivation and interest, giving them an explore independently (or at least less-dworld which they live. Finally, the idhelplessness can be minimized and the self-[4,11]. Figure 1 shows how the learneoccurs.

Fig. 1 – Generation of learned helplessness idea in d

Assistive Technologies have been provsevere disabled children a certain degree control by themselves [4]. This helps to taklearned helplessness, as shown in Figure 2.

Fig. 2 – Robotics support used to help blocking the leidea.

determine which ould allow them even more the ant.

k of interest of op the idea of selves as unable t external help. assivity and lack r she lives.

behavior, as said and dependents.

hild has a way to methods and

op [4,8]. Doing have a better opportunity to

ependently) the dea of learned -steam grows up ed helplessness

disabled children

viding to these of environment e out the idea of

earned helplessness

According to Swinth, Ansince their six-years old alreand cause and effect compute

Thus, assistive technologythese disabled children a uniqthey will interact with the encertain degree of control. opportunity to choose what thexperimenting only some ta[4]. Furthermore, robotic systhree dimensional object mreal situation compared withprovided by computer softwbe more realistic [4].

Several researches were dodisabled children could inter[4,10]. Nine children took pwith a disability and three wgroup of nine was less than was, basically, a computer connected with a small robmanipulator was used by thocan use that to bring to them

In this study, fifty percenwith cognitive age older thadisabled children interacted using it as a tool, to catch reach [4,10]. It was also oblinguistic level of the childrelevel [4,10].

Aftermost, this previous always focusing on explorenvironment by the children.

� Evaluate how severthe robotic manipula

� Define the relationsand the task comple

This research has shown year-old children use thenvironmental exploration [done analyzing how child baccess to a variety of movkeys, using the robotic in a d[4,2].

nson & Deitz [4,9], children eady have the ability to access er software by pressing a key.

y and robotics can provide toque opportunity to choose how nvironment and can also give a

So it is given to them the hey will experiment, instead of sks already designed for them stem also provide control over

manipulation, which is a more h the two dimensional graphics are, so the interaction can also

one to determine if very young ract with a robotic manipulator part of this research, being six without any. Everyone in this 38 months of age. The system to control and acquire data

botic manipulator [4,10]. The ose children as a tool, once they

objects [4,3].

nt of the disabled children (all an seven months) and all non-

with the robotic manipulator, an object that they could not

bserved that the cognitive and en were higher than the motor

research was complemented, ration and discovery of the The new goals were [4]:

re disabled children could use ator for exploration;ship between the keys pressed exity.

the way three years up to six e robotic manipulation for [4,2]. The same research was behave when they could have vements through one or more discovery and exploratory way

A complex progressive sequence task series was held and it was noticed an increase of cognitive development of those children. Consequently, as the task got more difficult, it was needed more pre-programmed keys to achieve the desired goal [4,10,3,8]. It was presented to the child how the manipulator moves using each key; showing the previous movement to them and encouraging them to use the system (press the keys) [4,6]. To maximize the results of those experiences dry macaroni inside a box and a glass were used. The task goals were:

1. First, the robotic manipulator should let the dry macaroni falling from the glass (where the dry macaroni were at first moment) by pressing a key just once;

2. Next, the child controls the robotic manipulator in order to fill the glass with the dry macaroni. The child gets the macaroni; put it inside the glass by letting it falls from the robotic manipulator, which is located above the glass;

3. Finally, and the more complex task of these test, is the overall sequence. First the child should get the macaroni from the box, put inside the glass and let it falls inside the box again. This must be done by using three keys.

As done in several works each one of these experiences were videotaped for further revision [4,10,2,6]. The research observations included the child’s action and its behavior during task accomplishment. For such example, it was observed how and who or what the child was looking during the tests. Behavioral signals were also included such as fear, happiness, if the child is smiling or crying, boredom or joy. It was registered before and during the use of robotic manipulator, so they could evaluate psychological aspects towards the experiment. Fortunately it was detected great happiness using the robotic manipulator, instead of fear [4,3].

It was also observed that child can respond longer times using robots instead of 2D computer software [4,2]. In those tests, it was also realized that the child could understand what each key can do by demonstrating its function previously [4,2]. High interest about the tests was observed when child looked towards the robot or the keys, proving they were interested on doing that [4,3]. This study complemented the previous works of the same authors, once it focused on the child’s understating about the system.

An important issue is the robot’s design. Tests done with autistic child has shown that the robot should be seen as a

toy and must be friendly and good-looking [13]. As related in [12] those child must feel comfortable and safe with the robots, instead of felling fear of them. Some properties as the robot speed and robot’s appearance should be carefully analyzed [12].

Although those articles cited previously ([12] and [13]) is about autistic children, the idea of the appearance of the robot can be applied to the disabled child, once the child must feel comfortable and safe with the robot. Other important thing is that tasks must be very joyful for those children, once boredom tasks often discourage them to use the system.

In autistic children the robots are used for better interaction, trying to take them out of their “own world” and bringing them to the “real world”. In the case of the disabled children the idea is to amplify their experiences inside the “real world”, increasing their contact with new and self-controlled experiences and consequently reducing the learned helplessness idea. So, some ideas can be used in both scenarios, although the two situations are completely different.

II. METHODOLOGY

For this research the main item used is a mobile robot with tweezers used for manipulation of objects. This robot will be commanded by the disabled children using their voluntary signals.

Several tasks will be done by those children, always focusing on environmental interaction, taking objects and finally putting them off on another place, finding “hidden” objects (actually an object among others with some degree of complexity to find), and drawing on a paper located under the robot.

All sessions will be videotaped for further revision (after the parents signing the Consentient Term, approved by UFES’ Ethical Commission). With the video, child’s actions (such as number and order of keys pressed necessary to complete, with success, the task), and behaviors during the experience can be evaluated better by a pedagogue.

III. HARDWARE AND SOFTWARE

The mobile robot used for this research is the POB-EYE, manufactured by POB TECHNOLOGY. It is a mobile robot with tweezers which allow using the robot as a manipulator robotic as well. For the task accomplishment it’s necessary to use an auxiliary device to catch the

myoeletric signals (EMG) together with signals coming from an accelerometer-baslast sensor can detect the child’s head inThose signals are further sent to a compBluetooth device. Three angles of view of thauxiliary device is shown in the figure 3. Tis the auxiliary device with the Bluetooth ch

Fig. 3 – Three angles of the robot and the circuit boarand inclination.

After sending the signal to the computanalyzes the signal and makes an interpretathe child wants. So the movement order tomoving) is sent to the robot. All the commuthrough Bluetooth devices.

While the computer is processing the sigrobot to move, the computer also makes a reach movement done, including time, pictuface in the moment of the movement andmovement. If there’s a correct sequence pthe report also shows if the movement is rigthis “correct sequence mode” the robot correct movements. All this data is imporevaluation by a pedagogue. Figure 4 showthe system.

Fig. 4 – Scheme of the system used to capture themovement the robot.

According to the child’s head movemendrive ahead, backwards, to the left and t

the inclination sed sensor. This nclination angle. puter through a he robot and the

The fourth frame hip.

rd to capture EMG

ter the software ation about what o move (or stop unication is done

gnal ordering the report recording

ure of the child’s d number of the pre programmed ght or wrong. In

only move in rtant for further

ws the scheme of

e signals and to

nt the robot can to the right and

other soft movements betweeTo make the tweezers’ moveEMG signal, so it is possibtweezers.

On the other hand, the acccontinuously and its value idefines where the child want

Limits values (maximum aare, actually, the highest anaxis. It is done throughout calculate the angle between t

Each value sent by the inthat will be used by the softchild wants to move the robvertical axis (front and back(left and right), which mepredominant. So if the incliand right the robot will drivethe system gets fully implemmovement with participationthat the robot will go in the d

The signals sent to the rwhich indicate what movcharacters are “w” for drivinright, “a” for driving to tbackwards. There is also thethat is represented by the cha

The same idea is done wthreshold trigger value to otweezers’ state.

Remembering that the robhas a pen hold by the tweezwith some training, to draw s

IV. COMPUT

The computer interface wand the child to achieve the gthe tests and allows addinauxiliary devices inside thsearching for reports. The mused to better conduct the tesdata to the test achievement.

This computer interface careas: main tasks, task rregistration, robot regisregistration.

en those four main movements. ements it is necessary to use the ble to switch the state of the

celerometer’s signal is received s converted to an angle which s the robot to go. and minimum) are defined and nd lowest inclination in each two axes, so it is possible to

them. nclination sensor is a number

ftware to understand where the bot. Firstly, it is analyzed the

k) and later the horizontal axis eans that the vertical axis is ination is diagonal being front e ahead, not to the right. When mented this will become a soft n of the two axis, which means diagonal direction. obots are, actually, characters

vement it should do. Those ng ahead, “d” for driving to the the left and “s” for driving e signal used to stop the robot aracter “t”. with EMG signal, which has a order the robot to switch the

bot is over a blank paper and ers, it is possible for the child, something on the paper.

TER INTERFACE

was done to help the evaluator goals. It resumes all the data of g robots, tasks, children and he database. It also allows main screen of the program is sts, once it shows all important

an be divided, basically in six registration, reports, children stration, auxiliary devices

The main screen is the task screen, whichimportant data for the test achievement. It control and access of all the other parts of thnecessary, before starting the test, to select a type of test, and a robot. It is also necescamera and the communication ports odevice and robot. The screen is shown in the

Fig. 5 – Main screen of the system.

The child’s name, robot and task is previin the software database. Such informimportant once it is used to make the report.

Connection ports of sensors and robot arthis part of the program. As said before, itselect a COMM port to allow communicatiand another one for the sensor. If the COMrobot is left blank or it is not possible to comthere is the possibility of doing the expervirtual robot, represented by a little yellowon the screen. On the other hand, if there virtual joystick in this screen can be usauxiliary device. The system scheme is show

Fig. 6 – Simplified flowchart of software op

This scheme considers that a child, a tescamera have been previously selected. If will show a dialogbox asking for choosing wOther important thing is that inside the blorobot” there is a process that will be explathe system “start” and “stop”, showed diagram, are, actually, the green and control’s groupbox.

h resume all the also allows the

he program. It is a child’s name,

ssary to select a of the auxiliary e figure 5.

.

iously registered mation is very . re defined inside t is necessary to ion for the robot MM port for the mmunicate with, rience using the

w circle, showed is no sensor the

sed as a virtual wed in figure 6.

peration.

st, a robot and a not, the system what is missing. ock “send to the ained further. In

in the above red buttons of

Inside this screen there is apart of the software allowscorrect movements and thmovement when it is correcregistered and can be used child has adapted well or not

V. AUTOMA

After doing the tests, a repchild’s name, the test name, tthe task, the number ofprogrammed sequence, if the

The list of movements dreport. This list contains thexcluding the movements Together with the number picture of the child that wshowing which movement th

If there is no right sequeblue. It there is a right sequegreen when the movement movement is wrong.

Following the arrows (anytext that says the direction there is the movement timmovement (the first movemmoment zero).

When the system is in “possible to calculate statistica

All this report is done evaluation by a pedagogue. to evaluate the emotional statheir face expression, such elements are, certainly, verysevere influence on the tests.

With some reports of thevaluate if they are or are nand if there is an improveterms. A report example is s

Fig. 7 – Report generated with thblu

a link to “right sequence”. This s the evaluator to choose the he robot only will do the ct. Wrong movements will be to evaluate statistically if the

t to the system.

ATIC REPORTS

port is generated, including the the robot name, the duration of f movements and the pre-ere is any. done is also presented in the he number of the movement, ordering the robot to stop. of the movement there is a

will be followed by an arrow he child did. ence, those arrows will appear ence, those arrows will appear is correct and red when the

y kind of them) there is a little of the movement and finally

me, compared with the first ment is considered to be at

“right movement mode” it is ally the system’s efficiency.

to further be possible and The photos are very important

ate of those children by looking as smiling or crying. Those

y important, once it can make ose children it is possible to not getting used to the system ement in behavioral-cognitive hown in figure 7.

he software. The images have been urred.

Supporting the main part of the programregistration part (of people, robots and auand the report screen.

Those registrations work similarly. Theyto a Compact SQL Database included inExpress Edition 2008 installation. Each reghas its own database.

VI. REGISTRATION AND DATA CO

In the child register some questions arechild’s name, birth date, which kind of dishas and the parent’s name. In the tasks reginame of task and a brief description is askethere is a unique identifier to ease the osearching.

In the registration of robots and auxiliamore information are asked, which are veryits use in the translation between the auxiliaand the robot signal. In other words, in thprograms it is asked which signal is recewanted to drive ahead, backwards, left odatabase knows previously which signal action.

At the same time, the robot registraticommand the robot should receive to do(such as driving ahead, backwards, left andof the tweezers mode). So the software caof the auxiliary device and “translate” iHence, using this system, virtually any acompatible with any robot can be used to peThis idea is shown in figure 8.

Fig. 8 – The correlation between the commands iequivalent items in different database

To illustrate this, a child would be ablerobot with any sensor. And on the other sensor he or she can control any robot (if it

m there are the uxiliary devices)

y are connected n Visual Studio gistration screen

ONVERSION

e asked, such as ability the child istration just the ed. In each case

organization and

ary robots some y important, due ary device signal hose parts of the eived when it is or right. So the represents each

ion asks which o certain action d right, changing atches the signal it to the robot. auxiliary device erform the tasks.

s done comparing es.

e to control the hand with one

is compatible).

VII. C

The main goal of this communication between the environment. It consequentland a better social interactioelements for human developm

Using the results of thepossible to change and creatmake those children inteenvironment, bringing themsteam.

All the system will be evchildren during a sequence o

REFE

[1] Thomas, RM (1992). Comparined). Belmont CA: Wadsworth Publi[2] Cook, Albert M.; Heng Max Development of a Robotic Device Who Have Severe Disabilities -engineering, vol. 10, Nº 3, Setembro[3] Cook, Albert M.; Meng, M.; DoFacilitate the Learning in Young Children – ART (Advanced Robotiof Alberta, Canada. [4] Cook, Albert M.; Howery, Kaexploration for very young childreInformation Resources), obtained iwww.dinf.ne.jp [5] Brinker RP & Lewis M (1982)infant: a process approach to assessmChildhood Spec. Educ., 2(2), 1-15. [6] Cook AM, Hoseit P, Liu KM, LArm System to Facilitate LearningIEEE Trans Bio. Med. Engr., BME-[7] Linder TW (1990). TransdFunctional Approach to Working wBrooks. [8] Scherzer AL & Tscharnuter I (1Cerebral Palsy. (2ª ed.) New York: M[9] Swinth Y, Anson D, Deitz J. (for infants and toddlers. American(11), 1031-1038. [10] Cook, Albert M.; Hoseit. Paul; an arm robotic system to facilitachildren - IEEE Transactions on bFevereiro 1988. [11] Todia B, Irvin LK, Singer GHesteem parent program. IN SingerDisability, and Empowerment. Toro[12] Werry I.P. and Dautenhahn K.the rehabilitation of autistic childRobotics Systems (SIRS’99), Coimb[13] Michaud F. and Clavet A. robotic toys for autistic children. PrEducation (ASEE’01) Albuquerque,

CONCLUSIONS

system is to increase the child and the external world or

ly will bring a better learning n as well, which are important ment. system’s reports, it will be te new tasks in order to try to eract even more with the

m more independence and self-

valuated with the help of those f tests they will perform.

ERENCES

ng Theories of Child Development (3ª shing.Q.; Gu, Jason J.; Howery, Kathy - for Facilitating Learning by Children

- IEEE Transactions on biomedical o 2002.ong, F.; Gu, J. - Using A Robot Arm to

and Mentally Development-Delayed ics and Teleoperation) Lab, University

athy – Robot-enhanced discovery and en with disabilities – DINF (Disabity in December 2, 1998. Site of DINF:

). Discovering the competent disabled ment and intervention. Topics in Early

ee RY & Zeteno CM: Using a Robotic g in Very Young Disabled Children, -35:132-137, 1998. disciplinary Play-based Assessment; with Young Children. Baltimore: PH

1990). Early Diagnosis and Therapy in Marcel Dekker.1993). Single-switch computer access n Journal of Occupation Therapy, 47

Liu, Ka Man; Lee, Ronald Y. - Using ate learning in very young disabled biomedical engineering, vol. 35, Nº 2,

HS, & Yovanoff P. (1993). The self-r GHS & Powers LE (eds) Families, onto: Paul H Brookes. Applying mobile robot technology to

dren. Proc. 7th Int. Symp. Intelligent bra, Portugal, July 1999;

Robotoy contest—designing mobile roc. American Society for Engineering , NM, 2001.

Educational Robotics as a Learning Aid for Disabled Children

Documents

Transcript of Educational Robotics as a Learning Aid for Disabled Children