<<

The NICE Pro ject Learning Together in a Virtual World

Andrew Johnson Maria Roussos Jason Leigh

Christina Vasilakis Craig Barnes Thomas Moher

Electronic Visualization Lab oratory Interactive Computing Environments Lab oratory

University of Illinois at Chicago Chicago IL USA

voice fax

niceiceeecsuicedu wwwiceeecsuicedu nice

Abstract VR and Education

Virtual realityenvironments designed for education typi

This paper describes the NICE project an immersive

cally fall into three categories networked textbased vir

learning environment for children implemented in the

tual environments desktop virtual simulations

CAVE and related multiuser VR tech

and immersive virtual environments

nologies The NICE project provides an engaging set

ting where children construct and cultivate simple virtual

ecosystems col labor ate via networks with other remotely

Textbased Virtual Worlds

located children and create stories from their interac

tions in the real and virtual world

Textbased virtual worlds commonly known as MUDs

and MOOs supp ort realtime interactive use by a large

numb er of users sharing the same virtual world from re

motely lo cated computers The advo cates of textual VR

cite two educational virtues of textdep endence the cre

Intro duction

ation of a communitycontext and the encouragementof

reading and writing Textbased VR has develop ed

The NICE pro ject is an eort to build Narrative

a strong connection to narrative and the area of Inter

based Immersive ConstructionistCollab orative Envi

active Fiction as participants can assume various iden

ronments for children Develop ed at the Interactive

tities develop situations cast in storylike formats and

Computing Environments Lab oratory and the Electronic

interact with ob jects and characters On the other

Visualization Lab oratory of the University of Illinois at

hand the lack of a sp ecic learning structure and di

Chicago NICE aims to create a virtual learning envi

rection mayleavemany participants oundering without

ronment that is based on current educational theories

sp ecic goals The lackof multisensory representation

of contructionism narrative and collab oration within a

may also render MUDs incomplete learning environments

motivating and engaging context

in and of themselves and textdep endence assumes liter

Designed to work in the CAVE tm and related

ate children

pro jectionbased VR hardware NICE allows groups of

children to learn together b oth in the same physical lo ca

Desktop Virtual Reality

tion as well as from remotely lo cated sites Additionally

awebbased comp onentallows other children to partici

Desktop virtual reality applications on p ersonal comput

pate from less exp ensive desktop hardware

ers allow users to walk through simulated environments

In section we rst briey describ e other currentwork created via readily available commercial Some

involving VR and education In sections and we slightly more exp ensive systems add p eripheral devices to

describ e the theoretical foundation and the implementa The VR pro jects provide a higher degree of interactivity

tion of the NICE pro ject In section we discuss evalua develop ed using these lowend technologies are limited

tion and then nally give our conclusions and plans for in size and complexity and lack immersive and interac

future work tive qualities Typical pro jects resemble D multimedia

simulations and mayinvolve the creation of simple mo d NICE

els and minimalist worlds which can b e exp erienced via

NICE Narrative Immersive Constructionist Collab o

walkthroughs

rativeEnvironments is an exploratory learning environ

mentdevelop ed for use bychildren b etween the ages of

and It is set on a virtual fantasy island where the

Immersive Virtual Environments

fairly large terrain includes a variety of spaces that in

vite exploration When the children arrive on the island

Immersive educational environments are b eing develop ed

they are greeted by the talking signp ost one of the many

using highend equipment They are consequently lim

intelligent guides or genies on the island which directs

ited to situations with sp ecial funding such as academic

them to several p oints of interest The children can climb

and researchenvironments Most of the pro jects are de

down a dormantvolcano to access the catacombs b eneath

velop ed esp ecially for headmounted display systems

the island lo ok for sh in the sea or help tend the garden

A large part of this educational research has b een

as shown in Figure

fo cused on science education as in the ScienceSpace

In the garden are several crates of seeds for the

pro jects Other pro jects such as the Virtual Real

children to cho ose from When they drop a seed on the

ity Roving Vehicle and summer camp programs in VR

ground the corresp onding vegetable b egins to grow The

for students at the Human Interface Technology Lab

children must make sure the plants are not to o close to

oratory at the UniversityofWashington fo cus on world

gether and that they get enough water and sunlight

building activities where students create the ob jects of

They can water the plants by pulling the raincloud over

their own virtual worlds using desktop D mo deling soft

top of them When a plant has had enough water it p ops

ware Although this gives the opp ortunity for students

up a little umbrella and the children can pull the cloud

to understand the pro cess involved in creating a virtual

away When the sunlightistoobringtheplants put on

setting the actual immersive exp erience is limited to a

their sunglasses when they feel crowded they moverest

short minute visit of the predesigned virtual worlds

lessly The children must keep the weeds from crowding

wers out the plants and decide what to do with the o

which o ccasionally blo om They can also shrink down to

the size of a mouse to get a closer view of the plants and

The Promise of VR for Education

to explore underneath the soil

VR oers promiseasalearning to ol as it allows access

The garden is p ersistent and continues to evolve so the

to the unreachable or the unrealizable it can provide

kids can come backandcheck on its progress at a later

multiple or alternative representations and it can help

time Multiple children can interact with the garden and

make abstractions more concrete VR is also touted

each other from remote sites Teachers or parents can

as a medium for interactive collab orative and engaged

participate either as memb ers of the groups or disguised

learning While highly resp onsiveinteractive virtual en

as genies This allows teachers to advise the children in

vironments can provide immediate feedbackand be ex

p erson or to direct parts of the activity from b ehind the

tremely motivating they may also foster automaticity

scenes They can also determine the pace at whichthe

thoughtlessness and lack of problem solving Telepres

world evolves they may cho ose to see the plants grow

the integration of VR and highsp eed network ence

very quickly or in the case of a scho ol pro ject extend

ing creates new p ossibilities for distance learning and

their growth over the p erio d of a semester

the exploration of so cial b ehavior amongst remotely lo

NICE is an outgrowth of two previous systems

cated learners sharing the same space However as the

CALVIN and the Graphical StoryWriter CALVIN

number of simultaneous users increase issues of control

is a networked collab orative environment for designing

and the structure of learning will emerge Virtual en

architectural spaces The Graphical Storywriter is

vironments are also valued as b eing extremely motivat

a shared workspace where young children can develop

ing esp ecially for students with nontraditional learning

and create structurally complete stories Building on el

styles but this engagement is due in part to the novelty

ements from these previous works we have created a

eect and not necessarily inherent in the design of the

prototyp e learning environment for young users which

learning environment

presents simplied ecological mo dels within a fantasy set

ting

The NICE pro ject which is seeking to explore the p o

tential of this promising to ol is describ ed next This highly graphical immersive virtual space has pri

Figure A child represented by an planting

carrots in the NICE garden

marily been designed for use in the CAVE a multi

p erson ro omsized virtual reality system develop ed at

the Electronic Visualization Lab oratory of the Univer

sity of Illinois at Chicago As the CAVE supp orts mul

Figure a small NICE story on the web

tiple simultaneous physical users a number of children

can participate in the learning activities at the same

time Since the CAVE library can supp ort heteroge

another pro duct of constructiveactivityinNICEisthe

neous VR display devices ImmersaDesktm Innity

narrative the stories formed by the kids that participate

Walltm BOOMtm shtank VR systems children

The narrative revolves arotnd tending the garden and

can join in the collab oration from a numb er of dierent

the reactions or decisions taken while interacting with

VR hardware platforms

the genies Every action in the environment adds to the

NICE fo cuses on bringing together concepts of con

story that is b eing continuously formed The narrative

structive group activity and narrative within the nexus

structure captures these interactions in the form of simple

of motivation and exploration In the following sections

sentences such as Amy pulls a cloud over the carrot

we will describ e the underlying p edagogical principles as

patchand waters it The tomato es complain that they

well as the technical asp ects of the system

are not getting enough water

This story sequence go es through a simple parser

which replaces some of the words with their iconic rep

Constructionism and Narrative

resentations and publishes it on a web page as shown in

Constructionism is an approach to learning that has re

Figure This gives the story a pictureb o ok lo ok which

ceived muchattention in the cir

the child can printout The children enjoy this pro cess

cles and is b eing applied to many current educational

esp ecially when they get to take home the written and

software pro jects It deals with the ways learners as

illustrated version of the narrative in either printed or

similate knowledge by engaging in selfdirected learning

electronic form

activities which are accomplished through constructive

tasks In NICE constructiveplayinvolves the plant

Collab oration

ing and tending of the garden

Constructionists b elieve that learning takes place when Equally imp ortant to the construction of ones knowl

engaged in the construction of a p ersonally meaningful edge is the exp erience gained by participation in group

artifact such as a piece of art a storyoraninteractive activity Collab oration is emphasized in our framework

computerized ob ject Aside from through the combination of collab orative learning across the garden itself

b oth virtual as well as physical communities

The network comp onentofNICEallows multiple net

worked participants to interact simultaneously with the

same virtual environment and each other These par

ticipants may be children or teachers or computer

controlled agents Multiple distributed NICE applica

tions running on separate VR systems are connected via

the central garden which guarantees consistency

within the shared virtual environment

Eachtracked participants presence in the virtual space

is established using an avatar with separate head b o dy

and hand as shown in Figure As each p ersons head

and hand are tracked this allows the environment to

transmit gestures between the participants such as the

no dding of the users head or the waving of the users

hand to the other participants As the child waves her

Figure Jim the avatar and Eddie working together

hand in the real world her avatar waves its hand in the

in the garden

virtual world These avatars have suciently detailed

representations so the participants can communicate no

tions of relative p osition to one another with phrases such

hardware devices

as it is b ehind you or turn to your left Ambient mi

NICEs networking comp onent allows clients other

crophones at each site allow the users to talk to each

than virtual realityinterfaces to participate For exam

other When a child sp eaks their avatars mouth moves

ple a recorder client can be connected to the network

to help dierentiate whichcharacter is sp eaking

that records all of the movements and interactions in the

Collab oration in NICE also includes the interaction b e

virtual environment This allows the session to be re

tween the kids and the genies which p opulate the virtual

played later during evaluation studies Monitoring clients

environment The genies op erate as interface and feed

can be connected to monitor the state of the NICE is

back mec hanisms disguised as helpful character guides

land and the state of the network Webbased clients can

in the virtual world They provide a way to replace

also connect and co op erate with the VR clients We will

menus and other traditional interface techniques with a

elab orate more on this in the next section

more transparent and eective metho d for use byyoung

children Many times when developing and testing the

system we use the WizardofOz prototyping technique

NICE on the web

where actual p eople disguised as genies simulate agent

behaviour The children maybeunaware of whichchar

Allowing webbased clients access to this immersive vir

acters in the environment are controlled by the computer

tual environment has several advantages for both the

and whichbyahuman op erator

users and the researchers in NICE

Virtual reality hardware is exp ensive and is far from

Collab oration also involves the so cial interaction

b eing integrated into the mo dern classro om NICE uses a

among children lo cated in the same physical space The

Silicon Graphics Onyx with two Innite Reality Engines

ories that emphasize the imp ortance of so cial interaction

for its graphics While the number of museum VR instal

in cognitivegrowth suggest that successful collab o

lations is growing the time that a child can sp end in an

rative learning involves more than the nal creation of

immersive virtual environment is still very limited The

textualized in a learning pro duct Learning that is con

webbased comp onentallows children to extend their in

a so cial setting mayinvolveverbal interaction collective

teraction with the virtual world b eyond the limited time

decision making conict resolutions p eer teaching and

they can sp end in the virtual environment itself We

other group learning situations characteristic of a class

have previously describ ed how the story of the childs

ro om setting With the use of VR technology that sup

exp eriences on the island can be published on the web

ports multiple users in the same physical space as well

but the web also allows the children to interact with the

as appropriate interaction techniques anumber of kids

virtual world

can participate in the learning activity at the same time

The web is also used to monitor the current state of the without having to take turns or wear heavy and intrusive

Figure Interacting with the garden through a java Figure Monitoring the garden from the NICE web

applet on the web page

Monitoring the Garden

garden and control various environmental parameters in

The NICE web page shown in Figure keeps track of

the environment The web also makes a convenientway

the current state of the p ersistent virtual environment

to distribute up dated mo dels for the environmenttore

including who is currently visiting the island and the

mote sites

current state of the garden This allows the users and

develop ers to monitor the state of the virtual environ

ment from a desktop computer with a webbrowser and

an connection The web is also used to moni

Collab orating in the Garden

tor the state of the network connections k eeping track

of more detailed information than is given to the casual

Students without access to a VR system liketheCAVE

users allowing us to further rene our networking proto

or Immersadesk can still participate in the collab ora

cols

tion through a web browser UsingaJava applet gure

children interacting with a twodimensional version

Controlling Garden Parameters

of NICE on the Internet can simultaneously share and

manipulate the same virtual space as the children in the

Visual languages have b een shown to b e ideal program

CAVE The users of the D environment use a tradi

ming languages for children b ecause they can map ab

tional mouse and icon interface to interact with the gar

stract concepts to pictorial elements that they are more

den but ha ve the same abilitytopick and plantasthe

familiar with Weenvision a visual language where

VR users do These desktop users see the virtual reality

the growth mo del for the garden and the b ehaviour of

users as D icons on their screen while the VR users see

the animal genies may be programmed by manipulat

the desktop users as D avatars in the space

ing iconic representations of data and op erators provide

immediate feedback and can b e learned quickly bythe

The webbased users can converse with the other vir

kids

tual and remote participants and co ordinate their actions

bytyping in the provided chat window a feature which The parameters in the garden server controlling the

resembles textbased virtual environments We are cur growth of the plants and vegetables on the island can b e

rently moving towards streaming audio from the virtual manipulated through this kind of interface Our current

world to the webbased clients to improve prototyp e is a twodimensional interface where the child

with the webbased clients trol the ecological by clicking and dragging icons to con

mo del can observe immediate feedback and eect in the

threedimensional VR environment

Downloading Mo dels

All of the mo dels used in NICE are VRML mo dels stored

on the NICE web server allowing us to leverage the ex

isting distributed server resources of the web EachVR

user keeps a lo cal cache of these mo dels When the en

vironment is rst launched the web server is checked

for anynewer mo dels and if any are found they can b e

immediately downloaded allowing the VR interface to

automatically upgrade itself

Similarlyeach user can create their own VRML avatar

and makethisavatar accessible from their lo cal web page

When they join in to the collab oration this avatar is

Figure Typical NICE networking connections all of

automatically downloaded to the other users This way

the interactive users are directly connected via TCPIP

each user is not constrained by the preexisting avatar

to the garden server running on the NICE web server A

mo dels but can create a very p ersonal avatar to represen t

CAVEaJava client and a VRML client are connected

themselves

directly to the router on the web server through UDP

A CAVE and an Immersadesk which share a common

multicast subnet are also participating in the collab ora

NICE Networking

tion through a lo cal router The session is b eing auto

matically recorded and two users with web browsers are

NICEs networking architecture was based on exp eri

watching

ences with several previous collab orative immersiveen

vironments The networking proto cols selected were tai

lored towards the characteristics of VR data and the

within a subnet but it is inconvenienttohave to set up

ability to easily enter and leave the environment from

multicast tunnels for casual long distance collab oration

anywhere on the internet

So a router is continually in op eration which broadcasts

The number of participants is limited only by band

packets b etween multicast and UDP connections Multi

width and latency of the network In practice we have

ple routers can b e chained b etween distant remote sites

hadasmanyassimultaneous participants spread over

allowing some clients to communicate via multicast while

and Asia Mul continents North America Europ e

others use UDP to share the same information Figure

tiple distributed NICE applications running on separate

shows a typical set of connections in NICE

VR systems are connected via the central garden server

The garden server and a UDP router reside on the

to guarantee consistency across all the separate environ

NICE public web site This allows any CAVE or web

ments The NICE garden is p ersistent as the garden

browser to join into NICE from anywhere on the inter

server is constantly running and will attempt to reestab

net NICE has p erformed robustly in transcontinental

lish lost connections in the event of failure This allows

collab orations at SIGGRAPH Sup ercomputing

users to casually join in the collab oration whenever they

SIGGRAPH and to Japan and the Netherlands at

wish The garden continues to evolve even if no users

Sup ercomputing

are present the plants continue to grow animals may

trytoeatthevegetables and weeds slowly takeover the

As we want to allow connections to the Java inter

garden crowding out the other plants

face even for users with K baud mo dems a certain

NICE uses a combination of reliable and unreliable amount of ltering is necessary to allow those users to

proto cols For the information from the garden server interact in realtime with the VR users Filtering is p er

whichcontrols plant growth picking and planting a re containing tracker formed at the router on the packets

liable TCPIP connection is used F information since this is the most abundant data and or the avatar po

loss of small amounts of this data will not greatly aect sition and orientation data multicast and or UDP is

the consistency of the world used Multicast is preferred when the collab oration is

User Studies Observed Results

Avarietyoftechniques were employed for gathering data

It is imp ortanttoinvestigate the educational ecacy of

b oth emb edded in the computer and external The ob

VR in sp ecic learning situations and broader learning

served results were collected byconverging the multiple

domains and to develop new rubrics of educational ef

pieces of data gathered through observation interviews

cacy that compare it to other approaches In practice

group ed according to an and questionnaires and were

however the assessmentofVR technology has b een fo

evaluation framework which was designed to take into

cused primarily on its usefulness for training rather than

consideration the unique asp ects of the medium This

its ecacy for supp orting learning in domains with a high

framework included guidelines for the study of the re

conceptual and so cial content

sults with resp ect to their technical orientational aec

The education world would argue that using pap er and

tive cognitive and p edagogical asp ects

p encil in the form of standardized tests is not an eec

Although the technical and physical issues concerning

tivewaytoevaluate a virtual learning exp erience As VR

the use of this new technology bychildren were not the

is a dynamic learning to ol evaluation should b e tightly

primary fo cus of these studies several general observa

coupled with the actual learning pro cess Following the

tions were made The children had little diculty learn

authentic assessment mo del learning in constructivist

ing how to use the interface technology and displayed

learning environments is directly related to its evalua

minimal instances of physical fatigue They did how

tion Moreover considering the immature nature of the

ever feel limited by the technology in terms of their

eld at this time it is imp ortant to apply multiple mea

ability to orient themselves and these technical limita

sures of learning and p erformance

tions hindered b oth usability and learning

It is generally argued that the aective factor encom

passing interaction and engagement is one of the most

imp ortantadvantages virtual reality has to oer to edu

Metho dology

cation Indeed during these studies all children were ex

tremely motivated and continued to b e engaged through

In the Spring of we observed a total of ftytwochil

out multiple sessions and after hours of participation

dren years of age while interacting with each other

Most imp ortantly however we observed that the chil

and the virtual en vironment Before their VR exp eri

drens engagementwas not limited only to increased mo

ence the children completed a set of pretest questions

tivation The amount of interactivity and engagement

as well as concept maps aimed at testing their ability

directly inuenced the outcomes related to the eective

to implement what they plan and orient themselves in

ness of the learning pro cess Learning was directly tied

the virtual world Two groups of children at twophys

to the childrens of engagement the ones actively

ically separate virtual reality systems were studied each

involved were more engaged and motivated to complete

time Each group included children one of whichwas

a task and consequen tly understo o d the mo del employed

chosen as the leader or co ordinator of the task The two

by the NICE garden The leaders the ones controlling

groups collab orated remotely represented bytheavatar

for the most part the interaction retained the most out of

of their leader to create and maintain a garden similar

the exp erience while the children that were less engaged

to the one they planned out b efore entering the virtual

did not understand the concepts as well While the ab ove

world In addition to the twoavatars sharing the same

suggest that nd graders can work eectively in VR it

virtual space an adult acting as teacher was disguised

is much harder to conclude whether anything conceptual

as another avatar and participated from a third VR sys

was learned during this exp erience Prestudy questions

tem The teachera vatar served multiple purp oses to

and interviews showed p ercent of the students had a

guide the children on their task to facilitate their dis

go o d understanding of environmental concepts while

course while in the environment and to be part of the

p ercent did in the p oststudy Of the p ercent p er

learning pro cess while observing and evaluating it In

centwere leaders of their various groups Our studies are

this way learning and assessment to ok place simultane

now fo cusing more on this cognitivearea

ously during the virtual activity Each session lasted for

an average of minutes and was followed byinterview Additionally our observations indicate that the use of

questions drawings and an essay completed by the chil teachers as avatars in the environmentwas p edagogically

dren Some observations from this study are given next eective esp ecially with regards to collab oration The

These studies are discussed in greater detail in teacher participating as another character in the world

acted as mentor from b ehind the curtains and help ed C Dede M Salzman and B R Loftin Maxwell

keep the children on task Nevertheless the teacher was World Learning Complex Scientic Concepts Via

in some instances still p erceived as an adult and had to Immersion in Virtual RealityInProceedings of Sec

sp end time attempting to keep order not unlike a real ond International Conference of the Learning Sci

classro om ences pages

These studies provide only preliminary observations in

C E Hughes and J M Moshell ExploreNet pages

the area of virtual reality learning They may serve how

The Virtual Reality Caseb o ok Van Nos

ever as a base for further exploration in the area of vir

trand Rheinhold New York

tual realityevaluation and address issues in the design of

virtual learning environments

J Leigh and A E Johnson Supp orting transcon

tinental collab orativework in p ersistent virtual en

IEEE Computer Graphics and Applica vironments

Conclusions and Future Work

tions

Although the cost and size of the CAVEoreven smaller

V Pantelidis Virtual Reality in the Classro om Ed

VR technologies prohibits their use in to days classro om

ucational Tecnology April

eorts are made to conduct studies at classro oms with

p ortable VR systems As the technology continuously

S Pap ert Mindstorms Children Computers and

decreases in cost and increases in p ower more museums

Powerful Ideas Basic Bo oks Inc New York

and cultural centers are b eginning to acquire multiuser

J Piaget To Understand is to Invent The Future

VR systems whichallow for manychildren to participate

of Education Grossman New York

in the exp erience at the same time

While wecontinue our studies we are using exp erience

M Roussos Issues in the design and evaluation of a

gained from the previous studies to improve the virtual

virtual reality learning environment Masters thesis

environment In the area of in terfaces we are rening

Electronic Visualization Lab oratory University of

the interface bydeveloping more natural mo des of inter

Illinois at Chicago

action that improve the direct manipulation of the envi

M Roussos A E Johnson J Leigh C R

ronment In the area of the garden itself we are work

Barnes C A Vasilakis and T G Moher The

ing on a scalable growth mo del where younger children

NICE Pro ject Narrative Immersive Construction

would see a simpler mo del while older children would

istCollab orativeEnvironments for Learning in Vir

see a more complex and realistic mo del In the area of

tual Reality In Proceedings of EDMEDIAED

collab orative learning we are investigating issues of self

TELECOM pages Virginia

representation nonverbal communication and the use

AACE

of VR as a way to enforce role dierentiation

NICE has also served as a valuable technological and

E SolowaySJackson J Klein et al Learning The

educational prototyp e environment for several other col

ory in Practice Case Studies of LearnerCentered

lab orative educational worlds we are currently creating

Design In Proceedings of ComputerHuman Inter

action pages New York ACM

Press

References

K E Steiner and T G Moher A Comparison

M Bricken Virtual Reality Learning Environments

of Verbal Interaction in Literal and Virtual Shared

Potentials and Challenges Computer Graphics

Learning Environments In Proceedings of ACM

CHI vol Interactive posters pages

M Bricken and C M Byrne Summer students in

VR a pilot study pages Virtual Reality

L S Vygotskii Mind in Society The Development

Applications and Explorations Academic Publish

of Higher Psychological Processes Harvard Univer

ers Professional

bridge MA sity Press Cam

A Bruckman and M Resnick The MediaMOO

this research is partially supp orted by NSF grantCDA

Pro ject Constructionism and Professional Commu

which includes supp ort from DARPA

nity Convergence