A Pilot Study of the Effectiveness of Augmented Reality to Enhance the Use of Remote Labs in Electrical Engineering Education

Abstract

Lab practices are an essential part of teaching in Engineering. However, traditional laboratory lessons developed in classroom labs (CL) must be adapted to teaching and learning strategies that go far beyond the common concept of e-learning, in the sense that completely virtualized distance education disconnects teachers and students from the real world, which can generate specific problems in laboratory classes. Current proposals of virtual labs (VL) and remote labs (RL) do not either cover new needs properly or contribute remarkable improvement to traditional labs—except that they favor distance training. Therefore, online teaching and learning in lab practices demand a further step beyond current VL and RL. This paper poses a new reality and new teaching/learning concepts in the field of lab practices in engineering. The developed augmented reality-based lab system (augmented remote lab, ARL) enables teachers and students to work remotely (Internet/intranet) in current CL, including virtual elements which interact with real ones. An educational experience was conducted to assess the developed ARL with the participation of a group of 10 teachers and another group of 20 students. Both groups have completed lab practices of the contents in the subjects Digital Systems and Robotics and Industrial Automation, which belong to the second year of the new degree in Electronic Engineering (adapted to the European Space for Higher Education). The labs were carried out by means of three different possibilities: CL, VL and ARL. After completion, both groups were asked to fill in some questionnaires aimed at measuring the improvement contributed by ARL relative to CL and VL. Except in some specific questions, the opinion of teachers and students was rather similar and positive regarding the use and possibilities of ARL. Although the results are still preliminary and need further study, seems to conclude that ARL remarkably improves the possibilities of current VL and RL. Furthermore, ARL can be concluded to allow further possibilities when used online than traditional laboratory lessons completed in CL.

Full text

A Pilot Study of the Effectiveness of Augmented Reality
to Enhance the Use of Remote Labs in Electrical Engineering
Education
A. Mejı
´as Borrero •J. M. Andu
´jar Ma
´rquez
Springer Science+Business Media, LLC 2011
Abstract Lab practices are an essential part of teaching
in Engineering. However, traditional laboratory lessons
developed in classroom labs (CL) must be adapted to
teaching and learning strategies that go far beyond the
common concept of e-learning, in the sense that completely
virtualized distance education disconnects teachers and
students from the real world, which can generate specific
problems in laboratory classes. Current proposals of virtual
labs (VL) and remote labs (RL) do not either cover new
needs properly or contribute remarkable improvement to
traditional labs—except that they favor distance training.
Therefore, online teaching and learning in lab practices
demand a further step beyond current VL and RL. This
paper poses a new reality and new teaching/learning con-
cepts in the field of lab practices in engineering. The
developed augmented reality-based lab system (augmented
remote lab, ARL) enables teachers and students to work
remotely (Internet/intranet) in current CL, including virtual
elements which interact with real ones. An educational
experience was conducted to assess the developed ARL
with the participation of a group of 10 teachers and another
group of 20 students. Both groups have completed lab
practices of the contents in the subjects Digital Systems and
Robotics and Industrial Automation, which belong to the
second year of the new degree in Electronic Engineering
(adapted to the European Space for Higher Education). The
labs were carried out by means of three different possi-
bilities: CL, VL and ARL. After completion, both groups
were asked to fill in some questionnaires aimed at mea-
suring the improvement contributed by ARL relative to CL
and VL. Except in some specific questions, the opinion of
teachers and students was rather similar and positive
regarding the use and possibilities of ARL. Although the
results are still preliminary and need further study, seems
to conclude that ARL remarkably improves the possibili-
ties of current VL and RL. Furthermore, ARL can be
concluded to allow further possibilities when used online
than traditional laboratory lessons completed in CL.
Keywords Robotics Digital design Augmented reality 
Virtual worlds Virtual scenarios Online education 
Augmented reality lab
Introduction
The main goal of this work is to achieve something that
until now seemed very difficult: Students can carry out labs
remotely on a real laboratory (not simulated or virtual),
also including virtual elements that interact with the real
elements of the laboratory, thus achieving an augmented
reality scenario. This opens up enormous possibilities for
students whose class attendance (for economic reasons,
work, distance, etc.) is problematic.
Following Andu
´jar and Mateo (2010), a VL simulates an
actual laboratory, is contained within one or more com-
puters and is equipped with content management and/or
learning capacities. RL is an environment (computer
application, VI, etc.) that allows acting on a real system
Electronic supplementary material The online version of this
article (doi:10.1007/s10956-011-9345-9) contains supplementary
material, which is available to authorized users.
A. Mejı
´as Borrero (&)J. M. Andu
´jar Ma
´rquez
Escuela Te
´cnica Superior de Ingenierı
´a, Universidad de Huelva,
Ctra. Palos de la Frontera, s/n, 21071 Palos de la Frontera,
Huelva
e-mail: mjias@uhu.es
J. M. Andu
´jar Ma
´rquez
e-mail: andujar@diesia.uhu.es
123
J Sci Educ Technol
DOI 10.1007/s10956-011-9345-9

remotely so as to teleoperate it, experiment and access data
through the net to obtain real measures. Of course, VL and
RL can integrate by performing supplementary but separate
functions (different and distinguishable).
ARL is a step further in online lab teaching and learning,
since—when endowed with all capacities for access, con-
tent and learning management available for VL and RL—it
is capable of connecting the real world of RL and the
virtual world of VL to set up scenarios where reality and
virtuality interact with one another, thus shaping an aug-
mented reality (AR).
AR—while incorporating virtual reality content—is a
distinct technology from virtual reality (VR) itself. VR is
isolated from reality and conforms to purely virtual sce-
narios. AR systems (Azuma 1997):
•Combine real content (usually observed through some
electronic device such as cameras and HMD displays)
and virtual computer-generated content, adequately
superimposed on the real content
•Are real-time interactive systems
•Must be registered in 3D space: the real space observed
by the user defines the context used to interact with and
represent real and virtual elements.
In short, AR supplements real-world perception and
interaction and allows the user to view a real environment
augmented with computer-generated 3D information. AR
applications, where virtual objects are aligned with and
superimposed onto the real world, enable the preservation
of the real user environment that provides a reference
frame for user actions, thus making human–computer
interaction more natural.
Between totally real and totally virtual situations there is
a continuum (Fig. 1), characterized by various mixtures of
virtual and real environments. In this mixed reality the
concept of a virtuality continuum (Milgram et al. 1994)
appears. This concept covers both AR and augmented
virtuality (AV), which are a mixture of the real and virtual
worlds. These intermediate points are also collectively
known as mixed reality. Figure 1shows a picture of one
real and one virtual environment, as well as both envi-
ronments combined by means of an AR application
developed by the authors of the present work that shall be
explained later on. AR can be observed to complete the
observation of the real world with virtual elements related
to the former. In this case, virtual objects are obstacles
found by the real minirobot in the left-hand figure and with
which it interacts. This is the key issue regarding AR:
interaction between reality and virtuality to shape a richer
‘‘reality’’ offering further possibilities.
Regarding Fig. 1, the route from reality to virtuality
(AR) can coincide in a specific point with the route from
virtuality to reality (AV). However, the term AR is usually
employed for any intermediate point between reality and
virtuality, since its current applications are closer to the
real than to the virtual world, among other reasons. AR
systems offer a clear advantage: the use of the real world.
Indeed, AR applications need not model every little detail
of reality; these details are already physically present
because they are real. It is only necessary to superimpose
those 3D virtual elements meaningful for the application
with which the user wants to interact. The user never loses
contact with the real world and, at the same time, can
interact with the superimposed virtual information.
AR is currently being introduced in new application
areas such as historical heritage reconstruction (Huang
et al. 2009), training of operators of industrial processes
(Schwald and De Laval 2003), system maintenance
(Henderson and Feiner 2009), or tourist visits to muse-
ums and other historic buildings (White et al. 2004),
among others.
The need for laboratory practices in engineering (of
course in other educational disciplines); to allow students
to acquire skills in solving real problems can present
logistical, economic and educational problems, including:
•Limited resources in the laboratories, both of software
and hardware.
•Real laboratory models are expensive. It is very
difficult to provide individualized material for each
student. Moreover, many universities will have prob-
lems to provide scaled-down industrial plants (in this
paper the concept of ‘‘plant’’ is that usual in control
engineering: the system which is to be measured and
controlled) for each working group (twenty students in
this case).
•Laboratory schedules must conform to the university
hours.
•The time available for each working group is always
insufficient, since the laboratories are shared between
different degrees and courses.
•Virtual laboratories typically have the added problem
of a lack of contact between the student and the
laboratory equipment.
These and other problems that arise, depending on each
individual institution’s situation, can be overcome using
the ARL system developed by the authors, which improves
real laboratory in various aspects among which include the
following:
•Real laboratory models can be expanded, reduced or be
modified. This allows a single plant to be used for
different experiments without having to modify the
physical environment. This can provide great savings in
cost and in preparation time for whoever is teaching the
lab sessions.
J Sci Educ Technol
123

•Laboratories with their plants and instruments can
operate 24 h a day.
•Laboratories can be used concurrently. A group of users
could be using a part of laboratory equipment remotely
by ARL, while another group might be physically in the
laboratory using other different equipment.
•Students interact remotely with real systems as if they
were physically in the laboratory, in front of the
equipment.
ARL is being tested in two required courses at the
second year of the new degree in Electronic Engineering
(EE), at the Higher Technical School of Engineering,
University of Huelva, European Union. The specific
courses are Digital Systems and Robotics and Industrial
Automation. The EE degree come under The European
Higher Education Area (EHEA), an initiative of the
Bologna process (Bologna process 2011) designed to create
more comparable, compatible and coherent higher educa-
tion systems in Europe (European Commission website
2009).
By means of AR, ARL allows the student to explore
learning experiences that may exceed those offered by
traditional laboratory classes, in the sense that it is possible
to develop virtual elements (only inside the computer) that
interact with the actual present in the lab, which allows to
configure an augmented reality. To illustrate the multiple
capabilities of this proposal, two very different practical
applications—the design of a digital control system based
on an FPGA (Field-Programmable Gate Array is an inte-
grated circuit designed to be configured by the customer or
designer after manufacturing; the FPGA configuration is
generally specified using a Hardware Description Lan-
guage, HDL) development board and the interaction of
remote real robots with virtual scenarios—are presented.
The development of these two practical ARL applications
has allowed to complete an educational experience in
which both teachers participating in the experience (TPE)
and students have taken part and whose results are shown
in this work.
To facilitate the reading of this work and due to the large
number of acronyms that contains, these are included in
Table 1.
Augmented Reality in Engineering Education,
an Overview
The academic world has also begun to introduce AR in
some academic disciplines, although its teaching applica-
tions are still minimal. The still-embryonic state of this
technology and its high development and use costs, as well
as its low presence in the everyday world, are amongst the
most important reasons why its still low level of
implementation.
Several European projects have designed and developed
innovative applications that integrate AR for educational
purposes, such as CREATE (Loscos et al. 2003) and
ARiSE (ARiSE Project 2009). These tools, based on 3D
presentations and user-interaction, facilitate science com-
prehension, as students can interact with virtual objects in
an augmented real environment and develop learning
experiences.
In another the educational use of AR (Liarokapis et al.
2004) puts forward an educational application for
mechanical engineering teaching that allows users to
interact with 3D content using web technology and AR-VR
techniques. Esteban et al. (2008) show AR for math
teaching, while (Kaufmann and Schmalstieg 2003)
describes a system for geometry teaching based on these
techniques.
An approach to the use of remote laboratories with AR
techniques is described in (Salzman et al. 2000), where
LabView
TM
is used to control an inverted pendulum in a
laboratory, although no monitoring or video-image location
Fig. 1 Reality-virtuality
continuum
J Sci Educ Technol
123

techniques are used, since the virtual image in this case is
derived from measurements on the real system.
Dormido et al. (2008) also contribute a development for
remote access to several types of experiments with
AR-supported visualization. These experiments include the
Heatflow system, which allows students exploring param-
eter estimation (identification) and controller tuning tech-
niques in delayed transport systems, as well as controlling
the position and speed of a DC motor.
Different e-learning systems using AR in the field of
Robotics have been developed in the last years for educa-
tional purposes. Albeit with reduced capacities still, some
representative examples are:
•UJI On Line Robot (Marı
´n et al. 2005): A complete
vision-based online robot system that allows control-
ling robots via web. Its interface is predictive: by means
of a 3D virtual environment endowed with AR
capabilities, the user can predict the results of the
actions before sending the command to the real robot.
•The ARITI system (ARITI 2010) also presents a display
interface enabling any person to remotely control a
robot via web. The Man–Machine Interface (MMI) is
based on the mixed reality concept, grouping VR and
AR, allowing easily perform of a task and description
of the desired environment transformation (to be
completed by the robot).
•Jara et al. (2008) present the development and imple-
mentation of RobUaLab. In this approach the real
information from the robot scenario is supplemented
with some virtually-generated data from the virtual
environment. Virtual projection is combined with the
current state from the remote laboratory taking current
IP camera setting and 3D environment into account.
This feature helps to handle the robot, providing further
information about its current and future situation.
All of these examples, but use AR techniques, none of
them allow the robot to interact with virtual elements. This
is precisely one of the successes of this work, which can be
seen later in the Lab # 2.
AR Techniques for Remote Lab Experimentation
One of the most evident advantages of remote labs is that
the process of preparing and carrying out the experiment is
very similar to that followed when physically in the lab.
Dormido (2002) lists the disadvantages of a remote testing
environment (e.g., the lack of physical contact with the
experiment can reduce the sense of realism). For this rea-
son, the AR-based method proposed in this paper is aimed
at giving the user the sensation that lab functions can be
handled just as they would be in the lab itself, thus
reducing possible discouragement due to the lack of
physical contact.
Another disadvantage listed in (Dormido 2002) is that
certain scientific/technological areas are unsuitable for
remote testing (e.g., chemical labs and the assembly of
combinational and sequential digital circuits in a digital
electronics lab). The first practical application developed in
this paper contradicts this assertion and shows how to
include digital design in those areas that could be incor-
porated into remote experimentation.
With respect to the requirements listed in (Dormido
2002), having an open and modular architecture is neces-
sary for the success of a remote experiment, as it allows
new components and exercises to be included with minimal
effort. AR techniques allow the same physical configura-
tion of lab equipment to be adapted for a wide range of
different experiments, since their respective equipment
needs are replaced by virtual elements that only appear if a
specific experiment requires them. In this sense, this pro-
posal presents important advantages, since the elements
Table 1 List of acronyms
Acronym Meaning
ANFIS Adaptive neuro-fuzzy inference system
AR Augmented reality
ARL Augmented remote lab
ARRL Augmented reality for remote laboratories
AV Augmented reality
CL Classroom lab
CMS Control management system
DAQ Data acquisition system
EE Electronic engineering
EHEA European higher education area
EPFL Ecole Polytechnique Fe
´de
´rale de Lausanne
FPGA Field-programmable gate array
HDL Hardware description language
HMD display Head-mounted display
IR Infrared light
LED Light-emitting diode
PID controller Proportional–integral–derivative controller
RL Remote lab
RS-232 Recommended standard 232
TCP/IP Internet protocol family: transmission control
protocol (TCP) and internet protocol (IP)
TPE Teachers participating in the experience
USB Universal serial bus
VHDL Hardware description language for very
high speed integrated circuit
VL Virtual lab
VR Virtual reality
J Sci Educ Technol
123

and virtual models connected to real equipment allow
many different experiments to be designed without altering
the real lab environment. If the remote laboratory is con-
sidered as an online educational service, this conception
dramatically reduces the down time required for physical
configuration modifications in the lab.
As mentioned above, AR has potential to augment
learning in many different knowledge fields as well as its
evident potential in teaching. This paper puts forward two
examples of a remote lab with AR techniques: the first one
is a practical application of a simple digital system based
on a development board with a FPGA. This implementa-
tion allows the remote use of the materials students nor-
mally use in classroom laboratories, particularly in the
design of digital systems. This educational experience is
focused on the interconnection of signals between virtual-
ity and reality so that, by means of AR, the real system is
augmented and improved by the virtual systems related to
it. Both systems (worlds), real and virtual, share signals and
communicate to one another creating synergy. Thus, the
final result is more complete and offers more possibilities
that obtained by real and virtual systems when acting
separately (see Fig. 1).
The second example has a completely different
approach: the remote implementation of a control system to
allow a real robot moving in a virtual scenario by avoiding
obstacles. The desired virtual scenario can be chosen pre-
viously. Besides, this scenario can be modified interac-
tively in real time, remotely and even when the real robot is
moving. This educational experience is aimed at making
the robot interact with the virtual scenario and learn to
move in it. Ultimately we aim at enabling students to train
robots in scenarios and situations that would be rather hard
to reproduce in lab.
The method proposed in this paper tries to meet stu-
dents’ needs to access the lab with a flexible schedule for
two kinds of labs, without the need of their having to come
to the university and with the same feelings as when being
physically present in the lab, while also adding new edu-
cational features.
Augmented Remote Laboratory
The ARL is a step above the VL and RL (Andu
´jar and Mateo
2010); the height of this step depends on the degree to which
it depends upon AR techniques. Access to the ARL is
enabled by an application that we have called ARRL (aug-
mented reality for remote laboratories). This application is
locally run on the user’s computer and grants access to the
remote lab via TCP/IP by means of AR techniques which
enable the interactive use of the lab equipment.
The tasks of the ARRL application can be summarized
as:
•Showing the student the remote device through a real-
time image taken by a remote camera located in the lab.
•Overlaying virtual elements (which enable the student
to interact with physically remote elements while
having the feeling of working directly on them)
correctly on this image. These elements must maintain
an appropriate position, scale and perspective in
relation to the received video image.
•Showing the results provided by the equipment. These
devices can be real (and thus seen in the video image)
or virtual (application-generated and properly placed
depending on the implemented design), with the
information being updated according to lab equip-
ment-provided data.
•The ARRL application must provide the student easy
access to experiment-related educational materials
provided by the teacher (e.g., the handout for the
experiment, tutorials, student manuals, etc). The access
to a content management system from the ARRL
application facilitates both the student’s access and the
teacher’s information distribution.
Figure 2shows the general structure of the developed
ARL system, where the following main elements are
distinguished:
a. Content Management System (CMS). This contains the
educational material that can be accessed from the
ARRL application run by the student through a menu
on the video image. Mambo CMS is used in the
prototypes.
b. Reservation manager and access control system. The
reservation manager ensures appropriate lab-resource
allocation, allowing students to reserve slots for their
use according to teacher-imposed restrictions. The
access control system uses the reservation data
provided by the reservation manager and enables the
necessary network resources.
c. Users (students). Students are asked to download from
the CMS the wording of the practice to be completed,
as well as the ARRL application which grants remote
access to the experiment and allows them to interact
with it so as to assess the result of their design.
d. Physical installation in the laboratory. The different
types of practices completed in the ARL share the
following two elements:
1. An IP camera. An IP Axis 211 camera is used in
the examples put forward next. It obtains the
original video image used by the ARRL applica-
tion to display the experiment. The ARRL
J Sci Educ Technol
123

application also superimposes the 3D virtual
elements with their correct position, size and
orientation.
2. A communications and access-control server is
used to provide access (from outside the Univer-
sity network) to the lab PCs that are directly
connected to the experiments.
3. PCs directly connected to the afore-mentioned
experiments. A PC can be connected to multiple
experiments by different resources (series ports,
USB ports, etc.)
e. Teacher. The teacher prepares the experiment that
students are to complete in the usual way. Once that
experiment suitability and feasibility is checked, the
teacher uploads the wording of the practice, technical
manuals and any other relevant information to the
CMS. Besides, the teacher also uploads the ARRL
application to the CMS for students to download and
run it in their PCs.
It must be noted finally a quality that is provided with
the ARL designed that might be able to pass unnoticed but
which is of critical importance. It refers to the power
supply of the equipment. All ARL equipments are powered
via a rack with TCP/IP connection. When the student
accessed remotely via the ARRL in order to log in ARL,
communications system automatically powers the neces-
sary equipment for the lab to perform (laboratory lighting,
power supplies, camera associated with the experiment,
instruments and equipment involved in the lab, etc.); sim-
ilarly, when the student finishes his remote session, the
communications system switches off automatically the
power of the material is no longer in use. In this way is
optimized the energy consumption of the whole installation
and preventing possible accidents.
The main difficulty in the development of the entire
methodology for ARL implementation is undoubtedly the
ARRL design, development and programming. Further-
more, ARRL is a highly time-consuming application and a
different version for each experiment could be necessary.
Few teachers will have the desire or time necessary to
develop this application. Therefore, the ultimate aim is to
create an ARRL application builder. In fact, we have already
developed a generator of ARRL applications for one of the
types of practices presented in Section ‘‘Examples of the
Application of the Developed Methodology’’ .
Examples of the Application of the Developed
Methodology
To illustrate the possibilities of the developed methodology
and the ARL, this section is devoted to explain the two
experiences completed for the first educational testing of
the developed system: The design of a digital sequential
control system using a development board with a FPGA (in
the subject Digital Systems) and the design of a control
system for a robot to avoid obstacles in a given scenario (in
the subject Robotics and Industrial Automation). Particu-
larly, it is about designing a control system and an inter-
active virtual environment in which a real robot moves like
a Braitenberg Vehicle, an agent (an artificial intelligence-
Fig. 2 General structure of the developed ARL system
J Sci Educ Technol
123

based controller) that allows the robot to move around
autonomously (Braitenberg 1984). For controller imple-
mentation, a fuzzy system obtained by adaptive techniques
is used in this case. The purpose of the lab is enabling the
robot to move without colliding with the user-created vir-
tual objects that make up the virtual scenario. The user can
choose the appropriate setting without modifying any
equipment in the remote laboratory. The user needs not to
put real obstacles around the robot, just load a virtual
world. The world he has imagined.
The use of virtual environments presents many possi-
bilities that would be unavailable in real environments.
Among them the following can be mentioned:
•The virtual elements that form the virtual scenario can
have virtually unlimited shapes and sizes. Moreover,
their shape can be fairly complex.
•They can be modified (size, shape, texture, position) by
the user remotely.
•Scenarios demand no maintenance in the lab, as they
are entirely composed of virtual elements.
•Virtual elements cannot deteriorate, break or hit
physical objects or equipment in the lab, or harm the
robot, either.
•The set of virtual objects can be uploaded from the
ARRL application and placed in the desired position
remotely.
•Remote trial assemblage and/or modification reduces
time consumption.
•Too complex and/or costly real scenarios (to be
physically implemented as lab models) can be
visualized.
•Likely scenarios are practically subject to the user’s
imagination.
Lab #1: Digital Sequential Control System: Mixing
and Level Control in a Tank
TPE and students are asked to complete an experiment with
a sequential system to control a tank (Fig. 3). The tank is to
be filled with a mixture of three different liquids, for which
purpose there are three valves (A, B and C) that control
liquid input, and four digital level detectors (n1, n2, n3 and
n4, where n1 represents the minimum and n4 the maximum
level). The tank is assumed to have a continuous internal
agitator that ensures uniform liquid mixing. The tank has
an output valve S than can be opened at any time for
mixture removal. Only when the reservoir level falls below
the minimum level n1, does a fill cycle take place: first with
liquid A up to level n2, then with liquid B up to level n3
and, finally, with liquid C to fill the tank (n4 level). The
liquid can then be removed from the tank. Note that in
order to simplify the experiment have not been considered
real world phenomena like stick slip in the valves, uneven
mixing, etc.
The teacher prepares the experiment with the standard
design method. In the example presented, the ISE WebPack
software (ISE Design Software 2010) is used to complete the
experiment ahead of time. This allows the teacher to identify
potential difficulties for TPE and students.
When the experiment has been shown to be entirely
feasible, the ARRL application is generated with the
builder. This application builder will examine the teacher’s
VHDL (Hardware Description Language for Very high
speed integrated circuit) project files, and collect and pro-
cess the necessary data (development board type, camera,
marker, IP configuration of server, etc.). With these data,
the application builder creates an ARRL application ready
to be placed in the CMS for TPE and students to download
and run it from their own PCs. Figure 4shows the aspect of
the developed generator of ARRL applications. When
available, the ARRL is uploaded to the CMS by the tea-
cher, thus making it available for TPE and students to
download and run it in their PCs.
These are the necessary phases to complete the experi-
ment, which ends with the programming of the remote
development board and the checking of its actual
operation:
Theoretical resolution: after downloading the experi-
ment text from the CMS, TPE and students must theoret-
ically obtain the state diagram of the digital system that
controls the tank. Figure 5shows the state diagram (Moore
machine) of this lab.
Writing of the VHDL module that describes the design,
simulation and programming of the development board.
The next step is obtaining a VHDL module that describes
circuit-design operation. The Xilinx ISE WebPack soft-
ware (freely downloadable from the manufacturer’s web-
site) was used for this.
Fig. 3 Tank to be controlled
J Sci Educ Technol
123

After syntax checking and design compilation, the TPE
and students must simulate the VHDL implementation. To
do this, from the above-mentioned software, TPE and
students write a test file in VHDL to represent the different
situations that may occur in the controlled system for
subsequent design simulation. The simulation is carried out
in the afore-mentioned design environment.
The TPE and students, whom have previously reserved a
post at the remote laboratory through the reservation
management system, program the development board
remotely from Xilinx software. Up to this point, no vari-
ation with respect to the usual steps necessary for system
design, simulation and board programming in the class-
room laboratory was introduced.
Design Verification in the Remote Development Board
This step clearly demonstrates the ARL possibilities.
Figure 6shows the general scheme of ARL to complete
this kind of practices. The following elements can be
observed:
a. Content Management System (CMS). It contains the
afore-mentioned educational material, which can be
accessed from the ARRL application that runs the user
through a menu on the video image.
b. Reservations manager and access control. The reser-
vations manager ensures appropriate allocation of
laboratory resources, allowing users to reserve slots
for their use as restrictions imposed by the teacher. The
access control system using the data provided by the
reservations manager enables the necessary network
resources.
c. ARRL application. This application is downloaded by
the user from the CMS. Is ready to be executed locally
within the runtime environment installed before, and
common to all applications.
d. Laboratory infrastructure. The training materials are
in laboratory as well as the necessary infrastructure for
access and remote viewing. A development board
connected to a PC (in experiments we use development
boards based on a Xilinx FPGA Spartan 3E series,
XC3S500E model with 232 IOBs and 10,000 logic
Fig. 4 ARRL application builder
J Sci Educ Technol
123

cells), a DAQ device, a PC used to remotely program
the board and an IP camera in front of the development
board. It is responsible for obtaining the original video
image, which is used by ARRL application to display
the board and superimpose the 3D virtual elements
with the correct position, size and orientation.
e. ARRL application builder.
Referring to Fig. 6: Once the ARRL application (c) has
been downloaded from the CMS (a) and runs locally, two
windows appear, one of which is the main application,
displaying a real-time video of the programmed remote
development board. In addition, the AR system is already
enabled on the actual image, appearing in this case as reset
button and four virtual switches (equivalent in this example
Fig. 5 State diagram (Moore
machine) of the system
Fig. 6 ARL general structure for lab #1
J Sci Educ Technol
123

to the level sensors) located exactly on their real counter-
parts on the board (Fig. 7a). These virtual elements can
now be manipulated with the mouse (also noting the
change in the position of the switch buttons) and the system
operation can be tested (the operation of valves A, B and
C) by the board LEDs (real), just as when physically in the
lab; this creates a synergy between the real and virtual
elements. Note that all the virtual elements are interacting
with the physical element (board) and vice versa, therefore
signals move from one element to another transparently to
the student. Figure 7b shows the original image captured
by the camera before virtual elements were put in place. As
it can be observed in Fig. 7b, the marker of the original
image is adequately covered by the ARRL application to
show the full board. The markers in the ARToolkitPlus
(2011) library were used.
The second ARRL application window is a console (not
shown in text mode as it contributes no clarity to the
explanation) where the application reports operational
events (e.g., switch performance) and possible errors.
It is also possible to observe the remote board and AR
elements overlapping with stereoscopic vision, which
requires using a stereoscopic camera, and configure the
application properly. We are evaluating whether this pos-
sibility brings benefits in this or other fields. Figure 8
shows a student performing the proposed experiment and
using stereo vision to visualize the development board and
a 3D model of the tank connected with the board. This
requires the use of electronic shutter glasses and appro-
priate display and graphics controller. We use a Samsung
SyncMaster 2233Z monitor, NVIDIA 3D Vision glasses and
aNVIDIA Quadro FX1800 graphics card.
Lab #2: Design of a Control System for a Robot
to Avoid Obstacles in a Scenario
TPE and students are asked to contribute a design for a
mobile robot to behave like a Braitenberg vehicle, but
using a fuzzy system obtained by adaptive techniques. The
purpose of the lab is that a real robot moves without col-
liding in a virtual scenario created by the user.
Before tackling the lab itself is necessary to perform a
series of preliminary considerations, which will value the
enormous possibilities of ARL developed.
Settings the ARL
For this lab, we use a robot Khepera II (K-team Cor-
poration 2011), a miniature mobile robot with similar
functionality to larger robots used for research and edu-
cational purposes. Khepera was originally designed as a
research and teaching tool for the Swiss Research Priority
Program at the Ecole Polytechnique Fe
´de
´rale de Lau-
sanne (EPFL). It allows real-world testing of simulation-
developed algorithms for trajectory planning, obstacle
avoidance, sensory information preprocessing, and
hypotheses on behavior processing, among others. Khe-
pera II (Fig. 9, left) is circular: a diameter of 68 mm at
its base and a height of 30 mm. It includes on-board
power supply and two independently-PID-controlled
motors. Moreover, it is equipped with eight IR (Infrared
light) sensors around the robot and positioned and
numbered as shown in Fig. 9(right). These sensors
embed both an IR emitter and a receiver.
Robots must fulfill these two conditions to be able to
interact with a virtual world:
Fig. 7 a ARRL Application main window. A drop-down menu can
be accessed in the upper left-hand corner. bOriginal image captured
by the camera
J Sci Educ Technol
123

1. The virtual world must use the same coordinate
system, which would have a real-world equivalent in
shape and size.
2. The robot must be able to detect the status of the
elements in the virtual world.
Both conditions can be solved through AR techniques.
Indeed, regarding the first condition, AR allows computer-
generated virtual elements to be referenced with a coor-
dinate system to the real world observed by a camera by
means of varied trackers: magnetic, optical, mixed, mark-
ers, etc. The marker used in this lab 2 is the same type as in
lab 1. Obviously, is a very cheap tracking method (just
print the marker), but the algorithm is highly time-con-
suming, because it must analyze each received video
frame, locating the marker and calculating its position in
the space observed by the camera, also using the camera’s
intrinsic parameters and distortion. The marker also has
some limitations, as it must be completely visible in the
video image; thus, lighting becomes a crucial factor and
must be as homogeneous as possible. The use of magnetic,
optical or hybrid trackers significantly improves object
tracking. It relieves the host CPU of tracking-calculus
operations, operated by outboard processors, which are
highly precise, but relatively costly too.
The second condition poses a very different problem:
the actual robot sensors (proximity sensors in Khepera II)
cannot detect a virtual world. This problem can be over-
come by creating virtual sensors which function as those of
the actual robot. Virtual sensors should be located on the
real robot in the same position and orientation as their real
counterparts. AR techniques also allow this overlapping of
virtual sensors. Thus, the ARRL application substitutes the
actual sensors with virtual sensors that can detect virtual
worlds.
Each sensor is modeled with three beams (Fig. 10) and,
when colliding with the virtual world, they return a value
proportional to the distance from the collision. An algo-
rithm within the ARRL application calculates the virtual
sensor output value in the same format as that provided by
the corresponding actual sensor in Khepera II. The virtual
body observed on the robot (Fig. 10) allows it to detect
collisions between real and virtual environments. In the
implementation of the ARRL application, the virtual body
and sensors are transparent, so that the user can only see
the real robot. For real-robot monitoring (tracking), a
marker allows software to calculate its position in AR
space.
Figure 11 shows the general structure of ARL for lab #2,
where it can be observed that it is a specific case of the
developed general ARL system shown in Fig. 2.
A socket server was developed to send commands to the
robot using TCP/IP communications. This server uses
multithreading techniques and is installed in the computer
directly connected to the robot. It receives and sends data
Fig. 8 A student performing the experiment using 3D vision
Fig. 9 Khepera II robot and
distribution of proximity
sensors
J Sci Educ Technol
123

to the RS-232 port that connects the robot (link 2 in
Fig. 12). The ARRL application constantly updates virtual
sensor readings and provides these data to the socket server
(link 1). When it receives a request to read the robot’s
proximity sensors, instead of sending this data, it returns
the last value provided by the ARRL application (link 3).
However, there is no reason to replace any other subsystem
in the real robot. Any other command sent from the robot
control program is sent immediately to the real robot, and
any response to this request is forwarded in turn to the
requesting control application (link 2). One of the advan-
tages of the developed methodology is that the robot is not
a simulation. Thus, all robot capabilities and responses
(except proximity sensors) are real. Link 4 in Fig. 12 shows
the real time video from an IP camera located in the lab to
the ARRL application. Another highly relevant advantage
is that, as it can be observed, the objective of the lab (which
is a function of each subject) and ARRL applications are
completely separated. Thus, in this case the control appli-
cation, objective of the lab, can be developed in any lan-
guage or development environment (C??, Matlab,
Python, LabView, etc.).
In the developed ARRL application (see bottom left of
Fig. 13), the user can select and move any virtual
obstacle with a simple mouse click. The selected element
can also be rotated on its axes; its size or aspect ratio can
also be altered, as show in the video (Online Resource
1). All these actions are unthinkable in a remote labo-
ratory using real elements in the area surrounding the
robot.
Fig. 10 Virtual items (virtual body and beams of virtual sensors)
superimposed on the real robot (visible to illustrate their position)
Fig. 11 ARL general structure for Lab #2
J Sci Educ Technol
123

Figure 13 shows a simple teacher-generated virtual
scenario (top left). This scenario is uploaded to CMS and
TPE and students use it to develop the mobile robot control
program. They can choose any of the obstacles and change
its position (for example the green wall at the top right),
rotate it (bottom left), or change its size (bottom right).
Once the desired transformation is performed, the object is
placed in the virtual scene with the new acquired charac-
teristics. The robot acts according to the TPE and student-
developed control program to try to avoid obstacles. These
changes can be made in real time, even while the robot
moves within the virtual environment. The Coin3D toolkit
(Coin3D 2011) is used for displaying 3D visualization.
Coin3D is built on OpenGL (OpenGL 2011) and uses scene
graph data structures to render 3D graphics in real-time. The
virtual scenario is defined in Open Inventor (Open Inventor
2011) file format, which can be interacted with by means of
Coin3D (this toolkit is fully compatible with SGI Open
Inventor 2.1).
Performance of the Lab
To perform this lab the TPE and student must obtain the
following from CMS:
Fig. 12 Diagram showing the
computer connected to the
robot, the ARRL application
and the communications server
(Tunneling Server)
J Sci Educ Technol
123

•The experiment text
•The toolbox for the Khepera II robot. This toolbox is
KMatlab (Piguet et al. 2010), but it has been modified
to allow access to the robot via TCP/IP, as shown in
Fig. 7. Matlab is used, as it is fairly common in
educational contexts and allows control system design
based on inverse model technique. It is probably the
simplest method for controller design based on the
adjustment of parameters from input–output data;
however, it is widely used to create behavior-based
control structures in mobile Robotics.
•The ARRL application, which includes two virtual
scenarios:
•The first scenario, used by TPE and students to
obtain 300 input data (proximity sensors) and
output data (speed applied to each wheel) from a
teacher-provided function which allows the robot to
evolve as a Braitenberg vehicle, storing the values
of proximity sensors and speed applied to each
wheel into a matrix.
•A second scenario, different to the first, where the
TPE and student must verify robot operation with
the designed control system, avoiding obstacles of
various kinds in different positions.
For control system development, TPE and students are
asked to produce two fuzzy systems to control two motors.
Fuzzy systems are then trained by ANFIS (Adaptive
Neuro-Fuzzy Inference System) (Jang et al. 1997) with the
data obtained while the robot evolved in the first virtual
scenario.
Finally, TPE and students must write a brief Matlab
script that controls the robot using the inferred fuzzy
systems. Logically, TPE and students can obtain many
results with this experiment: graphs representing the
training data and outputs provided by the inferred sys-
tems, errors, curves of relationship between the various
inputs (sensors) and output, testing the inferred systems
using a different set of vectors, verifying the results
using different membership functions (Gaussian, triangle,
etc.)
As explained above, control system design and TPE
and student-provided results are entirely performed with
Matlab (as in the lab with a real scenario around the
robot). The ARRL application runs only twice: to obtain
Fig. 13 Manipulating the virtual scenario from the ARRL application
J Sci Educ Technol
123

the training data that TPE and students use to train the
fuzzy systems with ANFIS and to visualize the robot
evolving in the second virtual scenario, while controlled
by the adjusted fuzzy systems.
First Tests as an Educational Tool
The developed work has a twofold purpose: learning and
teaching, hence to test it we have counted with teachers and
students. An educational experience was completed to
assess the developed ARL with the participation of TPE
and students. Both groups of users were asked to complete
lab practices by means of three different possibilities: CL,
VL and ARL. At the end of the educational experience,
both TPE and students filled in some questionnaires aimed
at assessing the improvement contributed by the ARL
relative to the remaining lab options available: CL and VL.
Each of both lab practices was completed by a group of 20
students (a whole group laboratory) chosen randomly
among 4 possible and another group of 10 teachers (the
number of 10 is casual, as that was the number of teachers
who voluntarily decided to participate in the experience),
and both practices are part of the contents in the subjects
Digital Systems and Robotics and Industrial Automation,
which belong to the second year of the new degree in
Electronic Engineering (adapted to the European Space for
Higher Education). Participating students were registered
in the afore-mentioned subjects and teachers did not give
lessons in these subjects, although they had enough training
experience to complete the educational experience. The
participating teachers and students belong to the Higher
Technical School at the University of Huelva (European
Union).
The lab practices completed by TPE and students
include the design of a digital sequential control system
using a development board with a FPGA (in the subject
Digital Systems) as well as the design of a control system
for a robot to avoid obstacles in a scenario (in the subject
Robotics and Industrial Automation). The developments
put forward in Section IV of this work were used in ARL.
The time that teachers and students have been available
to accomplish each lab is an hour and a half. This duration
was chosen because it is the usual in the classroom labs
schedules. The data included in Table 2refer to individuals
who perform the lab correctly in up to half an hour.
Table 2summarizes the completed practices and the
elements used. After lab practice completion, with enough
experience to compare the three types, both TPE and stu-
dents were asked to fill in a questionnaire containing a set
of questions on a 5-point Likert scale (1 = strongly disagree
and 5 = strongly agree) to assess ARL relative to the other
two proposals: CL and VL.
Tables 2and 3show that the use of ARL involves in
principle no inconvenience regarding lab practices, as the
number of students who are capable of completing the lab
practices properly is the same as when completed in
person (this is obviously a coincidence, yet it may also
indicate a trend). From our viewpoint, it may be due to
Table 2 Lab # 1: Design of a digital sequential control system using
a development board with a FPGA
Lab
type
Description Number and
ratio of
students that
solve the
practice
(max. time
allowed for
completion:
90 min)
(total no =
20)
Number and
ratio of TPE
that solve
the practice
(max. time
allowed for
completion:
90 min)
(total
no =10)
CL TPE and students design the
circuit and program the
development board in lab
16 80% 10 100%
VL TPE and students design and
simulate the circuit remotely
by means of the
manufacturer-provided
software. However, they have
no contact with the real lab,
as everything is virtual
17 85% 10 100%
ARL TPE and students design and
simulate remotely, record the
development board remotely
and check its proper operation
remotely on the real lab
16 80% 10 100%
Table 3 Lab # 2: Design of a control system for a robot to avoid
obstacles in a scenario
Lab
type
Description Number
and ratio
of
students
that
solve the
practice
(total no
= 20)
Number
and ratio
of TPE
that solve
the
practice
(total
no =10)
CL TPE and students design the control
system and test the robot in lab in a
scenario containing real obstacles
18 90% 10 100%
VL TPE and students complete the lab
practice by means of the
simulation software KiKS (see
Nilsson 2001)
13 65% 8 80%
ARL TPE and students design the control
system and ascertain its proper
operation remotely by means of the
ARL (real robot in a virtual
environment)
18 90% 10 100%
J Sci Educ Technol
123

the fact that ARL introduces no additional difficulty in lab
practices, but it is only a tool that allows us to interact
with the didactic material by means of AR and provides
remote support. However, design steps remain as in CL.
An analogous assessment would admit the results
obtained for TPE. The most striking result both for TPE
and students is in the case of VL in Table 3. The
exclusively virtual completion of the practice in the
robotic lab shows worse results. Undoubtedly this is due
to the additional effort involved by familiarizing oneself
with the simulator, as this step is not necessary in CL and
ARL.
The questionnaires are shown in Tables 4and 5which
contain the average responses of TPE and students on ARL
(Lab#1 and Lab#2). Both contain the same set of questions,
with two adaptations according to the lab (questions 1 and
11). Results show positive ARL general assessment
(question 14). Both TPE and students agree that applica-
tions have positive aspects regarding their graphic inter-
face, ease of use, installation and interactivity. Two sets of
answers are clearly different in both groups of users
(questions 5 and 6). They seem to show a trend in students
to think that theoretical concepts are learnt better through
practical applications. The authors of the present study
Table 4 Evaluation
questionnaire of the ARL
LAB # 1: Design of a digital
sequential control system using
a development board with a
FPGA
Question Description TPE Students
Mean SD Mean SD
1 Your level on digital systems design is high 2.8 0.63 2.05 1.28
2 The use of a graphic tool fosters the students’ motivation and interest 4.1 0.57 4.25 0.87
3 Putting the application into practice is feasible in the university
context
4.4 0.52 4.30 0.85
4 The application allows learning new theoretical concepts 3.3 0.95 3.6 1.26
5 The application allows consolidating theoretical concepts 3 0.82 4.5 0.84
6 Theoretical concepts can be learned through theoretical study alone 4.3 0.82 2.33 1.58
7 The software application has a clear and intuitive structure 4 0.94 3.94 1.17
8 The interface appearance is nice 3.7 1.06 3.75 1.27
9 The application is useful 3.9 1.10 3.80 1.17
10 The application is interactive 4 0.94 4.38 0.90
11 The use of virtual models is easy 4.1 1.20 4.19 0.95
12 The installation of the ARRL application is easy 3.9 1.19 4.3 0.88
13 The application facilitates theoretical-practical understanding 4 1.24 4.14 0.96
14 The overall assessment of the application is positive 4 0.82 4.2 0.89
Table 5 Evaluation
questionnaire of the ARL
Lab # 2: Design of a control
system for a robot to avoid
obstacles in a scenario
Question Description TPE Students
Mean SD Mean SD
1 Your level on robot control systems design is high 2.5 0.71 1.88 1.06
2 The use of a graphic tool fosters the students’ motivation and interest 4.2 0.63 4.16 0.84
3 Putting the application into practice is feasible in the university
context
4.6 0.52 4.08 1.05
4 The application allows learning new theoretical concepts 3.1 0.87 3.63 1.26
5 The application allows consolidating theoretical concepts 3.1 0.57 3.97 1.13
6 Theoretical concepts can be learned through theoretical study alone 4.1 0.87 3.05 1.67
7 The software application has a clear and intuitive structure 4 1.05 3.86 1.15
8 The interface appearance is nice 3.5 1.08 4.05 0.89
9 The application is useful 4 1.15 3.69 1.06
10 The application is interactive 4.8 0.42 4.52 0.74
11 The use of virtual scenarios is easy 4.9 0.31 4.38 0.73
12 The installation of the ARRL application is easy 4.1 0.99 4.22 0.93
13 The application facilitates theoretical-practical understanding 3.9 1.19 3.88 1.00
14 The overall assessment of the application is positive 4.2 0.92 4.02 0.84
J Sci Educ Technol
123

agree with this idea. However, TPE in Lab #1 do not seem
to agree to such an extent.
Regarding the influence of the use of ARL on learning in
the design of both VHDL-based digital systems and control
systems for robots, the high scores registered in questions
2-5 prove that these tools facilitate learning and are highly
useful in university contexts.
Finally, answers 7–13 show a high level of acceptance
by the users (both groups) of the developed system for ease
of installation, use, utility, etc. This opinion is even more
pronounced in Lab 2, perhaps for having a moving element
and make even more obvious (it is a visual perception) the
capabilities of the developed system.
Conclusions
Lab practices are an essential part of engineering education
and, therefore, efforts aimed at facilitating and bringing
these activities closer to students are numerous in different
universities.
The developments put forward in the present work are
aimed at contributing new means to remote lab practices
without neglecting the loss of realist sensation usually
involved by simulation. Bearing this purpose in mind, AR
is proposed as support technology in the education of
future engineers. We present fully operative applications.
All of them have been developed with free software tools,
thus allowing their diffusion among students with no use
restriction in educational contexts.
The possibilities contributed by AR are numerous. Thus,
the Higher Technical School at the University of Huelva
(European Union) is currently developing new applications
according to the ARL concept developed in this work.
We are aware that the educational study carried out is
very basic and may extend it more. In fact we think that the
developed system offers extraordinary possibilities in this
regard. However and according to the questionnaires,
the result of the educational experience proves that the
developed ARL may well be a powerful tool to improve the
remote teaching/learning binomial, particularly in a highly
difficult field such as remote lab practices in engineering
subjects. We expect that completion of the course
2011/2012 (June 2012) we will be able submit to the sci-
entific community a sophisticated educational study in
which the main argument will no longer be the developed
system (as in this work), but its statistical analysis as an
educational tool.
Future research shall focus on extending its applications
to other engineering fields (mechanic, chemistry, etc.)
through collaborations with other Departments at the
University of Huelva. On the other hand, the developed
ARL is currently being enriched regarding access control
and learning management.
References
Andu
´jar JM, Mateo TJ (2010) Design of virtual and/or remote
laboratories. A practical case. Revista Iberoamericana de
Automa
´tica e Informa
´tica Industrial (RIAI) 7(1):64–72
ARiSE Project, Augmented Reality in School Environments (2009)
http://www.arise-project.org/. Accessed 12 June 2011
ARITI Augmented reality interface for teleoperation via Internet
(2010) http://ariti.ibisc.univ-evry.fr/. Accessed 15 Jan 2010
ARToolkitPlus library (2011) http://studierstube.icg.tu-graz.ac.at/
handheld_ar/artoolkitplus.php. Accessed 20 June 2011
Azuma R (1997) A survey of augmented reality. PRESENCE:
Teleoperators Virtual Environ 6(4):355–385
Bologna process (2011) http://www.ond.vlaanderen.be/hogeronderwijs/
bologna/. Accessed 10 Jan 2011
Braitenberg V (1984) Vehicles: experiments in synthetic psychology.
The MIT Press, Massachusetts Institute of Technology, Cam-
bridge, MA
Coin3D (2011) 3D Graphics Developer Kit. http://www.coin3d.org/
Accessed 4 Feb 2011
Dormido S (2002) Control learning: present and future (Plenary
Lecture). IFAC 15th Triennial World Congress
Dormido S, Vargas H, Sa
´nchez J, Dormido R, Duro N, Dormido-Canto
S, Morilla F (2008) Developing and implementing virtual and
remote labs for control education: The UNED pilot experience.
Proceedings of the 17th world congress. The international
federation of automatic control, pp 8159–8164
Esteban P, Restrepo J, Trefftz H, Jaramillo JE, A
´lvarez N (2008)
Augmented reality: a space for the understanding of concepts of
calculus in several variables. http://www.eafit.edu.co/EafitCn/
Investigacion/Grupos/Ingenieria/RealidadVirtual/Realidad?
Virtual.htm. Accessed 12 June 2010
European Commission website (Education and Training) (2009).
http://ec.europa.eu/education/higher-education/doc1290_en.htm.
Accessed 10 Feb 2011
Henderson SJ, Feiner S (2009) Evaluating the benefits of augmented
reality for task localization in maintenance of an armored
personnel carrier turret. ISMAR-AMH. IEEE international
symposium on mixed and augmented reality—arts, media and
humanities, pp 135–144
Huang Y, Liu Y, Wang Y (2009) AR-View: an augmented reality
device for digital reconstruction of Yuangmingyuan. ISMAR-
AMH. IEEE international symposium on mixed and augmented
reality-arts, media and humanities, pp 3–7
ISE Design Suite Software Manuals and Help (UG681) (2010) Xilinx
Online documentation. http://www.xilinx.com/support/documentation/
sw_manuals/xilinx11/manuals.pdf. Accessed 16 Dec 2010
Jang JR, Sun C, Mizutani E (1997) ANFIS: adaptive neuro-fuzzy
inference system. In: Neuro-fuzzy and soft computing: a
computational approach to learning and machine intelligence,
1st edn. Prentice Hall, pp 335–368
Jara CA, Candelas A, Torres F (2008) An advanced interactive
interface for Robotics e-learning. Int J Online Eng 4(4):17–25
Kaufmann H, Schmalstieg D (2003) Mathematics and geometry
education with collaborative augmented reality. Comput Graph
27(3):339–345
K-team Corporation (2011) Khepera II user manual. http://ftp.k-team.
com/khepera/documentation/Kh2UserManual.pdf. Accessed 3 May
2011
J Sci Educ Technol
123

Liarokapis F, Mourkoussis N, White M, Darcy J, Sifniotis M, Petridis
P, Basu A, Lister PF (2004) Web3D and augmented reality to
support engineering education. World Trans Eng Technol Educ
UICEE 3(1):11–14
Loscos C, Widenfeld H, Roussou M, Meyer A, Tecchia F, Drettakis G,
Gallo E. Reche-Martinez A, Tsingos N, Chrysanthou Y, Robert L,
Bergamasco M, Dettori A, Soubra S (2003) The CREATE
Project: mixed reality for design, education, and cultural heritage
with a constructivist approach, the second ieee and acm interna-
tional symposium on mixed and augmented reality, ISMAR
Marı
´n R, Sanz PJ, Del Pobil AP (2005) The UJI online robot: an
educational and training experience. Auton Robots 15(3):283–297
Milgram P, Takemura H, Utsumi A, Kishino F (1994) Augmented
reality: a class of displays on the reality-virtuaity continuum.
Telemanipulators Telepresence Technologies SPIE 2351:
282–292
Nilsson T. 2001. KiKS is a Khepera Simulator. http://robotica.li2.
uchile.cl/el710/material/kiks_screen.pdf. Last accessed Sep 2010
Open Inventor (2011) http://oss.sgi.com/projects/inventor/ (last
accessed Apr 2011)
OpenGL (2011) The Industry’s Foundation for High Performance
Graphics. http://www.opengl.org/. Last accessed Jan 2011
Piguet Y, Legon S, Matlab K (2010) Matlab commands for Khepera.
K-Team Corporation. http://ftp.k-team.com/khepera/matlab/.
Accessed 14 Dec 2010
Salzman Ch, Gillet D, Huguenin P (2000) Remote experimentation:
improving user perception using augmented reality. NIWeek
2000:17–20
Schwald B, De Laval B (2003) An augmented reality system for
training and assistance to maintenance in the industrial context.
J WSGG, Int Conf Comput Graph, Vis Comput Vision 11(1)
White M, Mourkoussis N, Darcy J, Petridis P, Liarokapis F, Lister P,
Walczak K, Wojciechowski R, Cellary W, Chmielewski J, Staw-
niak M, Wiza W, Patel M, Stevenson J, Manley J, Giorgini F, Sayd
P, Gaspard F (2004) ARCO—an architecture for digitization,
management and presentation of virtual exhibitions. Computer
graphics international conference (CGI’04), pp 622–625
J Sci Educ Technol
123