Integrating Mixed-Reality Remote Experiments Into Virtual ... ? Integrating Mixed-Reality Remote Experiments Into Virtual Learning Environments Using Interchangeable ... reality remote experiments into virtual learning environments

Download Integrating Mixed-Reality Remote Experiments Into Virtual ... ? Integrating Mixed-Reality Remote Experiments Into Virtual Learning Environments Using Interchangeable ... reality remote experiments into virtual learning environments

Post on 01-Feb-2018

219 views

Category:

Documents

7 download

TRANSCRIPT

4776 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 56, NO. 12, DECEMBER 2009Integrating Mixed-Reality Remote Experiments IntoVirtual Learning Environments UsingInterchangeable ComponentsFrederico Menine Schaf and Carlos Eduardo PereiraAbstractThis paper presents a proposal to integrate mixed-reality remote experiments into virtual learning environments(VLEs) using the concept of interchangeable components, whichcan represent either real or virtual devices or software in in-dustrial automation systems. Combinations of real and virtualtechnical plants and automation systems are used in differentlearning scenarios for teaching control and automation concepts.Configurations of interchangeable components can be dynami-cally created via a VLE by configuring database parameters. Theproposed system includes a remote Web interface that follows athin client strategy and is designed to be compatible with Webbrowsers, including basic Java support. As the architecture thatsupports the integration of virtual and real components is locatedon the server side, remote students/users are only concerned withthe experiment and do not need to be aware of the system thatprovides this integration and flexibility. In the current version,interchangeable components are integrated via an Object Linkingand Embedding for Process Control interface, which is a widelyadopted standard in the control and automation area. The pro-posed approach also provides practical and theoretical supportfor experiments within a collaborative virtual environment. Thispaper includes a description of four experiments developed usingthe proposed environment and concepts.Index TermsCollaborative work, computer-aided engineer-ing, educational technology, learning systems.I. INTRODUCTIONENGINEERING education has become a crucial aspect formost countries since it has been recognized that skilledengineers are one of the main components for the developmentof innovative products and services, as well as for the opti-mization of production processes to ensure high productivityand quality. Considering education on control and automationsystems, a key issue is the reduction of the gap between clas-sical theoretical courses and real industrial practice. Hence, itis important to allow students to operate with devices, systems,and techniques that are as close as possible to those that theywill confront in industrial settings. Unfortunately, reproducingManuscript received August 15, 2007; revised June 18, 2009. First publishedJune 30, 2009; current version published November 6, 2009. This work wassupported in part by the Brazilian research agencies Coordenao de Aper-feioamento de Pessoal de Nvel Superior (CAPES), Financiadora de Estudos eProjetos (FINEP), and Conselho Nacional de Pesquisa (CNPq), by the Germanexchange agency Deutscher Akademischer Austauschdienst (DAAD), and bythe Servio Nacional de Aprendizagem Industrial (SENAI) in Brazil.The authors are with the Federal University of Rio Grande do Sul (UFRGS),Porto Alegre 90040-060, Brazil (e-mail: frederico.schaf@daad.alumni.de;cpereira@ece.ufrgs.br).Digital Object Identifier 10.1109/TIE.2009.2026369a real industrial plant in an academic environment is not an easytask. Industrial equipment is, in general, very expensive (bothin terms of acquisition and maintenance costs). Furthermore,safety constraints should also be taken into account. Suchfactors restrict the use of real industrial devices in academiclaboratories, which, in general, are then structured as small-scale experiments with little connection to industrial reality.Within this context, an industrial laboratory facility that isavailable via the Web and, therefore, accessible at flexible timesto a larger number of individuals helps to improve the overallcost effectiveness of such solution.The Internet growth has brought new paradigms and pos-sibilities in technological education. In particular, it allowsthe remote use of experimental facilities that can be used toillustrate concepts that are handled in a classroom and servesas an enabling and powerful technology for distance teaching.Through its worldwide connectivity, the Internet also allowshaving learning materials available to a much larger audience ofstudents. Web-accessible laboratories with remote experimentshave become an attractive economical solution for the increas-ing number of students [1]. They represent a second-best-of-being-there (SBBT) [2] solution for students and laboratorieswith expensive equipment. Following this trend, many institu-tions around the world have been engaged in the development ofWeb-based experimental settings. Systems aiming at teachingand research in several different areas have been proposed, suchas digital process control [3], proportionalintegralderivative(PID) control [4], embedded communication systems [5], su-pervisory control [6], robot and other systems teleoperation[7][9], and real-time video and voice applications. Mostly,these experiments utilize customized devices and software tomake small-scale textbook-like experiments remotely available.However, our experience has shown that the availabilityof remote experiments is not a sufficient condition to ensuresuccess in the learning process of control and automation engi-neers. Remote laboratory experiments offered as stand-alonesettings, without connection to adequate learning materials(explaining the topics that are to be learned in the experiment),usually lead students to the use of a trial-and-error strategy,which has a lower learning impact than originally expected.Moreover, remote laboratories are available 24/7 for a largeaudience of students, increasing the demand in the numberof faculty members and tutors that are necessary to provideonline guidance to students. To alleviate these problems,remote experiments can be integrated into virtual learning0278-0046/$26.00 2009 IEEESCHAF AND PEREIRA: INTEGRATING MIXED-REALITY EXPERIMENTS INTO VIRTUAL LEARNING 4777environments (VLEs) [10][12] that manage and provideguidance via learning materials before, during, and afterthe experimentation. This paper proposes such an integratedlearning environment on which mixed-reality laboratoryexperiments and student guidance tools are combined forcontrol and automation education. Mixed-reality experiments[13], on which simulated components can be combined to realequipment, are used to illustrate different learning situationsaccording to the knowledge level of remote students.The proposed environment has been developed with thecontext of the RExNet Consortium [14]. This Alfa II financedproject, funded by the European Community, had mainly threegoals: 1) to share; 2) to harmonize; and 3) to spread currentskills on remote experimentation. The first goal directly ad-dresses the essence of the ALFA program (financial support),namely, it calls for the cooperation among the consortiumpartners. Harmonization is a direct consequence of havinguniversities from distinct countries with different languages andcultures. Each university participating in the RExNet projectacts as a disseminating party within its own country, i.e.,spreading access to remote experiments to other surroundinguniversities. Results of the RExNet project can be seen in [10],[12], and [15][20].This paper is divided as follows. Section II describesDistance Education Environment of the Automation, Roboticsand Control Systems Group (GCAR-EAD), a VLE for con-trol and automation engineering education developed at ouruniversity. Sections III and IV describe experiment scenarios,which exploit different combinations of real and simulatedautomation-system modules. Section V compares the GCAR-EAD with other similar systems presented in the literature.Section VI discusses the results achieved when using theGCAR-EAD in undergraduate and graduate courses. Finally, inSection VII, conclusions are drawn and future work is discussed.II. GCAR-EADOur experience in applying remote experiments for controland automation started in 2002 with the construction of aremote laboratory with a Foundation Fieldbus (FF) pilot plant[20]. That system enabled students to learn PID tuning controltechniques and industrial communication protocols by workingwith a hypertext-based learning material and by remotely ac-cessing the pilot plant.A. FF Pilot PlantThe FF pilot plant consists of three interconnected tanks,four ac pumps (two of which are driven by vector-controldrivers), and five control valves to control the liquid flow inthe system. It is, therefore, possible to configure many differentsingle-variable and multivariable control loops, involving tanklevel and liquid flow as process variables, with various choicesof control inputs. All sensors and actuators in the plant arenetworked smart devices, which communicate using the FFprotocol.The use of a highly innovative industrial-oriented [21]technology brings a very strong perspective to research andteaching. Practical interest is raised in the students, as theexperiments become visually and conceptually closer to theprofessional practice. Also, a wider variety of concepts canbe illustrated, and several levels of difficulty and complexity/challenge can be explored.The pilot-plant architecture [20] is based on Object Linkingand Embedding for Process Control (OPC) and Web servicestechnologies to provide remote accessibility. Tools like theElipse SCADA1 software (supervisory system) and the Sysconsoftware and PCI302 FF network card2 (SMAR commercialproducts) were employed in the development and operation ofthe industrial pilot plant. The FF devices are responsible for thecontrol that is not affected by Internet connection delays.The communication between the supervisory and the Webserver is based on a client/server structure, which allows anybrowser (client) to access the screens of the supervisory system(server). The plant monitoring (visualization) is implementedby means of Java Applets. The user interacts with the browser,sending parameters to the plant via PHPMySQL functions.The information is processed by the supervisory system in theserver, and an answer is sent back to the user, closing the cycle.The experiment accessibility is controlled through an accesscontrol tool that has the following objectives: user validation,experiment scheduling, monitoring a sessions duration, para-meterization of experiments, and experiment initialization.B. GCAR-EAD: Embedding Remote Laboratories Into VLEsExperiences using the FF pilot plant showed that due to thefact that the learning material was loosely coupled with theremote experiment, students were not able to identify whichtopics to review in case they could not get the proposed experi-ments adequately done.To overcome those drawbacks, a system called GCAR-EADwas proposed, which supports mixed-reality remote experimen-tation. The GCAR-EAD has a more complex architecture (seeFig. 1) that additionally integrates learning-material manager(also called VLEs), educational materials, remote experimentswith mixed reality [13], [22], [23], interchangeable componentsstrategy [10], experiment analysis, and simple student guid-ance tools. The proposed architecture has five main modules:a learning-material manager, a student guidance system (orguide), experiment booking, experiment analysis (or analyzer),and experiment manager/interface. Each of these modules is re-sponsible for controlling a specific functionality of the GCAR-EAD. The interaction with each module is transparent, i.e.,students interact directly only with the VLE interface.A central database is the main communication channelamong modules (see Fig. 1). In the sequence, each of themodules is further described.1) Learning-Material Manager: This module containsall learning materials. All users are identified viausername/password, and, depending on the users category (incase of students, their knowledge level), distinct operationsare allowed when accessing an available learning material.Distinct learning modes are supported: active learning [24],1Elipse SCADA. See http://www.elipse.com.br.2SMAR PCI302; FF Process Control Interface. See http://www.smar.com/PDFs/Catalogues/PCI302CE.PDF.4778 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 56, NO. 12, DECEMBER 2009Fig. 1. GCAR-EAD high-level architecture.[25], distributed learning [26], and team learning [27]. Active-learning skills are justified since, via environment interactions,students can self-learn (or self-teach). Distributed-learningskill is obviously linked to the spatial flexibility characteristicoffered by VLEs Web accessibility. The most important skillis, however, related to collaborative interaction, i.e., studentteams (or users, in general) may work together, increasing theknowledge transfer in a common environment. In traditionalcourses, it is assumed that teachers were the only source ofknowledge, and they centralized all courses information. Thecollaborative learning skill is mostly associated with the socialconstructionist pedagogic line [28].The learning-material manager plays the most meaningfulrole in the GCAR-EAD architecture since this system wasdeveloped to supply an educational (learning) environmentattached to remote experiments for the students.2) Experiment Booking Module: This module is responsi-ble for controlling students access to experiments. Since realexperiments are not replicable with no cost, booking systemsare necessary to organize the use of the real equipment (orentire real experiments) by students. Only signed VLE userscan book/access experiments and select the available time slotsfor running their experiments in a calendar-type Web page.3) Experiment Interface/Manager Module: The experimentmanager provides a link among the remote experiments andthe VLE and must ensure that the right remote experimentinterface is available according to VLE setup parameters. Thatmeans that the experiment manager receives from the VLE areference to the experiment to be executed and constructsthe experiment, providing also a Java Applet interface for datavisualization.This module is also responsible for implementing the inter-changeable components strategy [10] by linking and combiningreal and virtual components in a learning scenario.4) Experiment Analysis: The experiment analysis modulecomprises tools to evaluate the results of a conducted experi-ment and determine some metrics derived from the experimentresults (for instance, in the PID tuning experiment, usual met-rics are overshoot, rise time, settling time, etc.). The experimentdata are supplied by the experiment manager and the VLE in aform of reports or are directly stored in the central database.Experiment feedback characteristics are also stored in thecentral database and are available for further visualizationand/or manipulation by the VLE and other modules.5) Student Guidance Module: The last module of theGCAR-EAD architecture is responsible for providing studentguidance, which means that it receives as input the metrics gen-erated by the experiment analysis module and has to determinewhether students have met the goals defined by tutors/teachers.If not, this module has to indicate learning materials to bereviewed by the students. Therefore, this module acts as a sim-ple tutoring system integrated in the VLE. The guidance musttake into account all information stored in the database relatedto the student (student level, previous experiments performed,and visited learning materials). The tutor can also analyzereports (history data) of completed quizzes and VLE-proposedactivities like assignments.C. VLE Integration With Mixed Reality SupportingInterchangeable ComponentsWhile the remote access of real laboratory equipment hasseveral advantages, there are also some issues to be consideredfor teaching control and automation concepts.1) The number of students/student groups working simulta-neously is equal to the number of physical experimentsavailable.2) Real systems with a slow dynamic lead to long waitingtimes.3) Interlocking systems have to be carefully developed toavoid that students may damage components via im-proper actuation.Such drawbacks can be overcome by employing simulatedcomponents. Simulations, although sometimes unrealistic, havesome advantages that can be explored in different learningscenarios. One of the main advantages of using executable sim-ulation models is that they can be easily replicated. Students canthen simultaneously use multiple copies (replicas) of the samesimulation simultaneously. The simulation replicas, instead ofreal experiments, do not imply on more equipment becausereplicas only consume more computer memory and processorpower. Another advantage of using simulation is that studentscan speed up slow dynamic systems for quick visualizationusing simulation models (e.g., heating a real tank can takehours, while simulation speeds up to seconds). Safety concernsinvolving simulation-variable limits are not as important as inreal experiments since the models cannot be damaged.By analyzing the pros and cons of real versus simulatedexperiments, one can see that, in some sense, they are com-plementary so that a combination of both possibilities seemsinteresting. The strategy called interchangeable componentshas been developed to allow the combination of both real andSCHAF AND PEREIRA: INTEGRATING MIXED-REALITY EXPERIMENTS INTO VIRTUAL LEARNING 4779virtual components [10]. This supports the definition of a vari-ety of learning scenarios. For instance, real automation equip-ment such as programmable logic controllers could be remotelyprogrammed by students, and these controllers would then con-trol a simulated industrial plant. Additionally, simulated plantscan be used to evaluate the robustness of control algorithmswhen the technical plant presents an unexpected behavior. Onthe other hand, simulated automation systems can be useful inshowing step-by-step execution of industrial controllers.Based on our experience in teaching control and automationcourses, three different learning scenarios can be identified.1) Fully simulated. This kind of experiment setup illustratesan experiment abstraction where simple and ideal simu-lation models (without perturbations and other real-worldcharacteristics) are employed. Although simulation mod-els are not necessarily simple or without perturbations,for didactic issues, the implementation of simple modelsis more adequate in early stages of experiment learningprocess.2) Mixed simulated/real components scenarios. This config-uration can be used, for instance, in the interaction be-tween a simulated controller and a real plant to elucidatehow acquired data from the real plant vary from the idealmodel, and this can cause instability in the controllerprogrammed logic; consequently, some precautions mustbe addressed in the simulated equipment to treat thatinstability. When dealing with a real controller and asimulated plant (hardware-in-the-loop), some problemsalso occur in the delay of the control logic since thecontroller cannot instantaneously process the acquireddata (commonly, the controller cycle time is responsiblefor this delay).3) Scenario with real components. This experiment sce-nario is the typical implementation of remote laborato-ries where SBBT [2] is implemented, and students canperform experimentation using real components and ob-serve how theory applies into practical applications. Here,nonlinear behavior, perturbations, physical constraints,communication delays, etc., affect the experiment, and allthese real-life limitations can be visualized.D. VLE Integration With Tutoring SystemsThe proposed VLE integration with tutoring systems is re-sponsible for every GCAR-EAD interaction feedback. Tutoringsystems are dependent on several other tools or modules. Eacharchitecture module stores information in the central databasecontributing to the tutoring-system feedback compilation.Basically, an integrated tutoring system gives two kinds offeedback: 1) allows remote experiment configuration accordingto the user (student) level, i.e., students with no previouslyrecorded interaction with the experiment should start with basicexperiments, whereas more advanced students can directly goto more complex experiments; and 2) infers, based on the as-sessment of the experiment performed by a student, the requiredlearning material to be worked.The first type of feedback compilation searches in the centraldatabase for previously performed experiments and visitedlearning materials for a given student or student group. Basedon these data, it decides which type of learning scenario canbe accessed. The second feedback type also takes into accountthe reports generated by the experiment analyzer to suggest aspecific didactic material to the student.The experiment analyzer plays a meaningful role in theexperiment-driven tutoring-system feedback. There are twoproposed types of experiment analyses: 1) For dynamic ex-periments, the result of the analyses (evaluation) is mostlycomputed offline, i.e., after the experiment has been concluded,control metrics like overshoot and rise time are calculated.2) On the other hand, discrete experiment based on logic controlcan be evaluated in execution time since the digital I/Os canbe tested while the experiment is running. The first type iscalled postruntime analysis, whereas the other is called runtimeanalysis, but both produce reports that are stored in the centraldatabase.The postruntime-analysis tool compares the results of theexperiments performed by students with the requirementsspecified in the experiment assignment. The runtime analysisincludes some testing codes in the experiments to verify if spe-cific conditions are satisfied. Although the evaluation is madein the execution time, the suggestion is also postexperiment.Stored experiment messages (experiment reports) are read bythe tutoring system, and learning-material suggestion is done incase the experiments goals were not reached.III. GCAR-EAD IMPLEMENTATIONThis section will describe some applications developed usingthe GCAR-EAD architecture presented in this paper. Most ofthe applications were developed within the scope of the RExNetProject [14].The learning-material manager module offers a user-friendlyinterface with the free open-source course-management systemcalled Moodle [28]. Moodle is a widely used software packagedesigned with social constructionist pedagogical principles tohelp educators create effective online learning communitieswith many collaborative learning tools.OPC-Data Access (OPC-DA) standardized interface is em-ployed to allow configuration of scenarios without building fourdifferent and distinct experiments (which include three learningscenarios). The OPC interface provides a common, simple, andreusable interface to the interchangeable components strategyimplementation. Basically, the same interface is used for stu-dents interaction with real and simulated components (and,therefore, they are considered interchangeable). Each of thecomponents (real/simulated plant and controllers) must havean OPC interface, so that automatic selection of the inter-changeable components can occur; consequently, configurationof scenarios can be automatized. Therefore, OPC works as aglue among different applications (in this case, components).The experiment booking, student guidance, and experimentanalysis modules are implemented in PHP code. A SCADAsoftware, named Elipse SCADA, is used as the experimentmanager and is responsible for providing connections with thereal/simulated equipment and for supplying Java Applets asexperiment visualization interfaces.4780 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 56, NO. 12, DECEMBER 2009Fig. 2. (Left) FF pilot plant experiment. (Right) Thermal plant experiment.All developed remote experiments (case-studies imple-mentations) follow common software architecture: ApacheWeb-server software, MySQL as a database interface manager,OPC-DA for experiment-level communication interface, andthe ISaGRAF as a simulation tool.IV. DEVELOPED REMOTE EXPERIMENTSFour major courses for control and automation educationwere developed in the GCAR-EAD environment: 1) PID con-trollers; 2) IEC 61131 Standard for programming industrialautomation applications; 3) FF communication protocol; and4) remote experiment course. Other related courses are indevelopment, e.g., a complete course on control systems theory.The PID course contains several learning materials relatedto the design of PID controllers, such as PID tuning methods,stability analysis, and system performance analysis. This coursemakes extensive use of remote experiments, aiming to providestudents with some hands-on examples.The second course introduces the IEC 61131 Standard, awidely adopted standard for programming industrial applica-tions. Students are required to program industrial applicationsusing the IEC 61131 languages (sequential flowcharts, ladderdiagrams, instruction list, function block diagrams, and struc-tured text).The FF communication-protocol course introduces conceptsof industrial communication protocols in general and includesa specific discussion on the FF protocol.The last course contains didactic materials referring to theconstruction and use of the remote experiments developed andintegrated in the GCAR-EAD.A. Enhanced FF Pilot PlantThe first experiment was developed using an FF pilot plant[10], [20], which has been used as a remote laboratory forseveral years for teaching PID controller theory, in particular,PID parameter tuning (see Fig. 2). The original remote lab-oratory was integrated into the GCAR-EAD (see Fig. 3) byincluding a new learning material on PID control theory as wellas incorporating an experiment analysis tool to calculate controlmetrics from the results of the performed experiments. In thisexperiment, students have to control the water level in two tanksby acting on pumps and valves.Fig. 3. Snapshot of the computer screen viewed by remote students whenperforming the experiment on the FF pilot plant in GCAR-EAD environment.Note that the water level could also be monitored by an online video feedback.The interchangeable components strategy is implemented inusing the script language of the SCADA software (Elipse).Moodle-embedded PHP programs write scenario parameters sothat the scripts configure the experiment, setting up links andconnections between selected components.B. Thermal PlantThe second experiment uses a simple thermal plant [5] builtwith a PID industrial controller and simple electronic equip-ment to illustrate temperature-control techniques and the use ofindustrial controllers (see Fig. 2). The experiment consists ofa thermal resistance and a PTC100 thermal sensor connectedto an industrial controller N1100 from the Brazilian companyNovus. This controller manipulates the electric current thatflows in the resistance via pulsewidth-modulation switchingand, thus, affecting the temperature of the resistance. Thisindustrial controller offers a PID control technique. Studentsmust, based on open-loop trails, identify and control the tem-perature (in closed loop) of the experiment by setting PIDparameters given temperature set points.C. Mixed-Reality Electropneumatics WorkbenchThe third experiment, used in vocational training, was de-veloped in collaboration with researchers from the Universityof Bremen and the University of Berlin in Germany [13],members of the RExNet consortium. In this case, the so-calledderiveSERVER, developed in Germany, is integrated via anOPC interface to the GCAR-EAD environment [12].The system provides mixed-reality experiments through theuse of hyperbonds, which allow tight coupling between phys-ical and virtual phenomena [13]. Physical signals such as airpressure and electric potential are converted into binary logicinformation and vice versa. Hyperbonds are based on the bond-graph theory [29], which provides a unified view on differentsystems using the notion of effort and flow.The system software architecture is entirely based onclientserver modules allowing great modularity and mobil-ity (system distribution). The hyperbond software interfacewas modified from the original German project also currentlysupporting, besides original hardware connectivity, OPC andparallel/serial communication interfaces.SCHAF AND PEREIRA: INTEGRATING MIXED-REALITY EXPERIMENTS INTO VIRTUAL LEARNING 4781Fig. 4. Mixed-reality workbench for mechatronics vocational training atSENAI-RS.Owing to the clientserver architecture of the hyperbond tothe experiment manager, i.e., Real Object Manager, studentscan use local hardware attached to a simple parallel PC portto interact with the experiment and also to local OPC serverswith any kind of compatible equipment.The new hyperbond interfaces offer remote hardware orsimulator interaction using any of the interfaces, i.e., studentscan have their own hardware/software (which are compliantwith the new interfaces) remotely interacting with the system.Didactic materials related to this experiment, in Portugueseand English versions, are divided into four major courses:1) devices manual; 2) use of the deriveSERVER system;3) experiment guide; and 4) basic electropneumatics.To focus the system utilization, task-oriented experimentscenarios were developed so that students have a well-specifiedgoal (see Fig. 4).D. Simulated Bottle Production PlantThe fourth experiment is a virtual experiment, whose be-havior is modeled in the ISaGRAF software using IEC 61131programming languages. This experiment provides a very di-dactic and reusable experiment that can be combined with theother experiments presented and has a built-in analysis tool,integrated into the simulation model, which can check thebehavior of automation systems developed by the students thatcontrol the process.V. RELATED WORKSAlthough there are several interesting related works pub-lished in the literature, we could not identify any solutionintegrating all concepts incorporated into GCAR-EAD.Similar FF pilot-plant experiments were encountered, such asCoupled Tank Apparatus WebLab [30] and Automatic ControlTelelab (ACT) [31].Several projects have employed and tested remote-experiment networks like LabNet [32], CyberLab [33], RemoteWiring and Measurement Laboratory [34], digital-signal-processing-based Remote Control Laboratory [35], DigitalElectronics Education and Design Suite (DEEDS) [36], andNetwork Control System Laboratory [37].The Solar Energy e-Learning Lab from Michaelides et al. [11]has an integrated learning system with several learning materi-als and quizzes to identify student-understanding level. First,the student must pass several theory tests so that the systemgrants remote-experimentation access. Despite these qualities,the system also does not offer experiment feedback.The Remote Lab System [38] is a much more elaboratee-learning environment, where a framework that addressesbandwidth efficiency for the remote monitoring and controlof course laboratories is employed. This framework providesinstructor-friendly remote monitoring of the laboratory, ef-fective evaluation, and grading methodology, and, differentfrom others, also supports effective help from the teacher. Theteachers help connection allows audio transfer and remotecontrolling of the students computer, which helps in case ofdoubt resolution.The Distance Internet-based Embedded System Experimen-tal Laboratory (DIESEL) [39] is a learning-environment sys-tem for remote experimentation that allows students to accessembedded hardware ranging from microcontrollers to real-timeoperating systems. This implementation also offers autonomousconfiguration of the experiment with intelligent user support,i.e., provides a semiautomated intelligent helping tool based ona case-based reasoning system that is able to monitor and infercontext-aware assistance to users on demand. Therefore, thissystem highlights the research pointed to autonomous help oruser support that a remote-learning environment can offer.The Power Electronics E-Learning Platform proposed in [40]consists of a reconfigurable power-electronics testbed, a Web-based distance laboratory, and a user interactive e-learningplatform. The reconfigurable power-electronics testbed can beconfigured via a Web-based interface to construct a wide varietyof converters and inverters. This way, this system offers muchmore flexible experiments.Another excellent example of flexible experiment configu-ration is the deriveSERVER system proposed in [13]. Unlikeothers, this employs mixed-reality techniques and is integratedin the VLE with collaborative and distributed learning methods.This solution, however, has no specific experiment goals orexperiment learning feedback. Thus, the user (student) neitheris guided nor receives any analyzed results of the performedexperiment.The ACT proposed in [31] offers not only controller para-meterizations but also MATLAB Simulink models to describeand characterize the controller logic. This interesting approachis very useful in the experiment configuration. Thus, studentscan design their own experiment controller in a much moreflexible way.VI. EXPERIENCES USING GCAR-EADThe GCAR-EAD is currently being used in undergraduateelectrical engineering lectures at the Federal University of RioGrande do Sul (UFRGS). So far, experience in adopting thesystem can be considered as very successful, either in termsof comparing the performance of students using the tool withothers that are not using it or based on students evaluation.A custom quiz was developed to evaluate system qualities andfaults according to this class of students. Results indicate that4782 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 56, NO. 12, DECEMBER 2009the majority of the students have enjoyed the environment and,in particular, the possibility of combining both real and virtualexperiments. The quiz also indicates that the time flexibility andthe collaborative environment are the most important character-istics according to the students opinion. Thus, obtained resultshave been very positive. In particular, one can see that studentsmotivation has increased. Logging data analysis shows thatwhereas some students access the remote experiments late atnight, others prefer to work early in the morning; that meanseach one can define his or her preferable working/learning time.The system is being continuously tested and improved withmany suggestions.As an example, the use of the GCAR-EAD system in an un-dergraduate lecture on Control Systems in the second semesterof 2007, which has proposed a blended learning method,combining remote virtual lessons with traditional classes, hasconsiderably increased in comparison with those in the previousyears (around 90% of students approved, 23% more than in theprevious semester with the same instructor).VII. CONCLUSION AND FUTURE WORKThis paper has presented a virtual Web-accessible learn-ing architecture. The GCAR-EAD environment was developedbased on the proposed architecture, which allows an integra-tion of mixed-reality experiments with VLEs, introduces theconcept of interchangeable components, includes experimentanalysis tools, and provides student guidance through the learn-ing material.Whereas the proposed environment has proven to be veryuseful for control and automation education, there are stillsome challenges to be faced. The synchronization in the timingbehavior of the virtual and real equipment is dependent onthe communication delays in the network infrastructure. In thecurrent implementation, this delay is around 2 s for the wholecommunication between the client and the end actuators. Whilethis is not meaningful for technical plants with slow dynamics(as is the case in the selected experiments), it has to be im-proved. Of course, there is a tradeoff in having geographicallydistributed applications and the higher communication timesthat are required.The proposed environment can also be used for collaborativeengineering since experiments can be distributed into severalsites, and several students (users) can interact using the sameenvironment.Although the work of Ma and Nickerson [41] presents aliterature review with skeptic research on the long-term debateof technology as the enabler of students learning, it is widelybelieved that collaborative experiences enabled by remote ex-periments associated with virtual environments are powerfuldrivers of cognitive processes and can significantly enhancelearning efficiency. The benefits of collaborative learning arewidely researched and advocated throughout the literature [42].Regardless of the varying theoretical emphasis in different ap-proaches on collaborative learning (e.g., social constructivism),research clearly indicates that, in many (not all) cases, studentslearn more effectively through collaborative interaction withothers. This motivates the preparation of remote laboratoriesfor collaborative learning and their use in distributed teachingscenarios with simulation tools, hands-on laboratories, andpractical workshops.There is a strong demand for research that seeks to createsuch blended learning, where collaborative remote laborato-ries can play a significant role. Emphasis on collaboration addsnew technical requirements to the design of remote laboratories.Mixed-reality concepts support learning environments withremote laboratories and distributed workspaces. The bidirec-tional telecooperation functionality allows students to use theInternet for collaborative engineering. The presented environ-ment allows groups of students/technicians (or even employees)at remote locations to take part in the same training usingthe same equipment (either simulated or real). The users areable to work in a collaborative way to solve problems andexplore solutions to proposed problems. This kind of interactionprovides a systematic support for skilled workers and engineers.The presented research can be used as an appropriate and cost-effective tool to support collaborative engineering.The hyperbond concept allows integration of different remoteequipment or simulators providing common environments toremote collaboration of experiments. This is especially impor-tant to cost-effective remote experimentation since real hard-ware must not be duplicated.While software may be designed to achieve closer socialties or specific deliverables, it is hard to support collaborationwithout also enabling relationships to form and to supporta social interaction without some kind of shared coauthoredworks. For this purpose, a new environment interface is un-der development. This new interface will employ play ethic(methods) applied to work-turning activities that employ com-puters for a more comfortable experience. This is commonlyreferred to as gamelike interface. A 3-D representation is alsoin study aiming to display more realistic (virtual) worlds to itsusers. In this manner, future works indicate new trends to theGCAR-EAD that will be further developed to become a com-plete computer-supported collaborative environment.ACKNOWLEDGMENTThe authors would like to thank their RExNet partners fortheir fruitful collaboration.REFERENCES[1] M. Cooper, The challenge of practical work in a euniversityReal,virtual and remote experiments, in Proc. Inform. Soc. Technol. Conf.,2000, pp. 3440.[2] B. Atkan, C. Bohus, L. A. Crowl, and M. H. Shor, Distance learningapplied to control engineering laboratories, IEEE Trans. Educ., vol. 39,no. 3, pp. 320326, Aug. 1996.[3] L. L. Bello, O. Mirabella, A. Raucea, and L. Capetta, ENEL Pilot: Froma research testbed to a virtual educational laboratory, IEEE Trans. Ind.Electron., vol. 56, no. 12, pp. 48444853, Dec. 2009.[4] C. Batur, Q. Ma, K. Larson, and N. Kettenbauer, Remote tuning of aPID position controller via Internet, in Proc. Amer. Control Conf., 2000,pp. 44034406.[5] C. Schmid and A. Ali, A web-based system for control engineeringeducation, in Proc. Amer. Control Conf., 2000, vol. 5, pp. 34633467.[6] J.-S. Lee and P.-L. Hsu, Remote supervisory control of the human-in-the-loop system by using Petri nets and Java, IEEE Trans. Ind. Electron.,vol. 50, no. 3, pp. 431439, Jun. 2003.[7] R. Marn, P. J. Sanz, P. Nebot, and R. A. Wirz, Multimodal inter-face to control a robot arm via the web: A case study on remoteSCHAF AND PEREIRA: INTEGRATING MIXED-REALITY EXPERIMENTS INTO VIRTUAL LEARNING 4783programming, IEEE Trans. Ind. Electron., vol. 52, no. 6, pp. 15061520,Dec. 2005.[8] L. Huijun and S. Aiguo, Virtual-environment modeling and correctionfor force-reflecting teleoperation with time delay, IEEE Trans. Ind. Elec-tron., vol. 54, no. 2, pp. 12271233, Apr. 2007.[9] A. Balestrino, A. Caiti, and E. Crisostomi, From remote experiments toweb-based learning objects: An advanced telelaboratory for robotics andcontrol systems, IEEE Trans. Ind. Electron., vol. 56, no. 12, pp. 48174825, Dec. 2009.[10] F. M. Schaf and C. E. Pereira, PID controller remote tuning experimentwith learning environment integration, in Proc. 12th IFAC Symp. Inform.Control Probl. Manuf., 2006, vol. 2, pp. 753758.[11] I. Michaelides, P. Elefthreiou, and D. Mller, A remotely accessible solarenergy laboratoryA distributed learning experience, in Proc. Int. Conf.Remote Eng. Virtual Instrum., 2004.[12] F. M. Schaf, C. E. Pereira, A. C. Assis, C. L. Reichert, F. Campana, andI. Krahkeche, Collaborative learning environment using distributedmixed reality experiment for teaching mechatronics, in Proc. 8th IFACSymp. Cost Oriented Autom., 2007.[13] F. W. Bruns and H.-H. Erbe, Mixed reality with hyper-bonds: A meansfor remote labs, in Proc. IFAC 11th Symp. Inform. Control Probl. Manuf.,2004, pp. 5568.[14] G. R. Alves, J. M. Ferreira, D. Mller, H.-H. Erbe, N. Hine, J. B. M. Alves,C. E. Pereira, L. Chiang, O. Herrera, and L. E. Sucar, Remote experi-mentation networkYielding an inter-university peer-to-peer e-service,in Proc. 10th IEEE Int. Conf. Emerging Technol. Factory Autom., 2005,pp. 10231030.[15] J. B. Silva, G. R. Alves, and J. B. Alves, Remote experimentation inte-grating research, education and industrial application, in Proc. 8th IFACSymp. Cost Oriented Autom., 2007.[16] Y. Hernndez, J. Noguez, and L. E. Sucar, Incorporating an affec-tive model to an intelligent tutor for mobile robotics, in Proc. 36thASEE/IEEE Frontiers Educ. Conf., 2006, pp. T2F22T2F27.[17] H.-H. Erbe and D. Mller, Distributed work environments for collab-orative engineering, in Proc. 7th Int. Conf. Inform. Technol. BalancedAutom. Syst. Manuf. Serv., 2006, pp. 46.[18] O. Herrera, G. R. Alves, D. Fuller, and R. G. Aldunate, Remote labs ex-periments: Opening possibilities for distance learning engineering fields,in Proc. 19th IFIP World Comput. Congr., 2006, pp. 321325.[19] J. M. M. Ferreira, A. M. Cardoso, and M. G. D. Gericota, An integratedframework to support remote IEEE 1149.1/1149.4 design for test experi-ments, Int. J. Online Eng., vol. 2, 2006.[20] R. P. Zeilmann, J. M. G. Silva, A. S. Bazanella, and C. E. Pereira, Web-based control experiment on a foundation fieldbus pilot plant, in Proc.IFAC Int. Conf. Fieldbus Syst. Appl., 2003, pp. 325330.[21] L. Costas-Prez, D. Lago, J. Faria, and J. J. Rodriguez-Andina, Opti-mization of an industrial sensor and data acquisition laboratory throughtime sharing and remote access, IEEE Trans. Ind. Electron., vol. 55,no. 6, pp. 23972404, Jun. 2008.[22] P. Milgram, Merging real and virtual worlds, in Proc. IMAGINA, 1995,pp. 218230.[23] M. Billinghurst and H. Kato, Collaborative mixed reality, in Proc. Int.Symp. Mixed Reality, 1999, pp. 261284.[24] K. Watson, Utilization of active and cooperative learning in EE courses:Three classes and the results, in Proc. Conf. Frontiers Educ., 1995, vol. 2,pp. 16.[25] K. A. Smith, The craft of teaching cooperative learning: An active learn-ing strategy, in Proc. Conf. Frontiers Educ., 1989, pp. 188193.[26] M. Auer, A. Pester, D. Ursutiu, and C. Samoila, Distributed virtual andremote labs in engineering, in Proc. IEEE Int. Conf. Ind. Technol., 2003,vol. 2, pp. 12081213.[27] N. Faltin, A. Bhne, J. Tuttas, and B. Wagner, Distributed team-learning in an Internet-assisted laboratory, in Proc. Int. Conf. Eng. Educ.,Manchester, U.K., 2002, pp. 15.[28] M. Dougiamas and P. Taylor, Moodle: Using learning communities tocreate an open source course management system, in Proc. World Conf.Educational Multimedia, Hypermedia Telecommun., 1995, pp. 171178.[29] H. M. Paynter, Analysis and Design of Engineering Systems. Cambridge,MA: MIT Press, 1961.[30] C. C. Ko, B. M. Chen, J. Chen, Y. Zhuang, and K. C. Tan, Developmentof a web-based control experiment for a coupled tank apparatus, IEEETrans. Educ., vol. 44, no. 1, pp. 7686, Feb. 2001.[31] M. Casini, D. Prattichizzo, and A. Vicino, The automatic control telelab:A Web-based technology for distance learning, IEEE Control Syst. Mag.,vol. 24, no. 3, pp. 3644, Jun. 2004.[32] F. Davoli, P. Maryni, M. Perrando, and S. A. Zappatore, General frame-work for networked multimedia applications enabling access to laboratoryequipment: The LabNet project experience, in Proc. IEEE Int. Conf.Inform. Technol.: Coding Comput., 2001, pp. 359365.[33] R. Haugom, O. K. Solbjrg, K. Y. Pettersen, and T. I. Eikaas, A simula-tion game for nonlinear control theory education, Model. Identif. Control,vol. 28, no. 2, pp. 4551, Apr. 2007.[34] J. A. Asumandu, R. Tanner, J. Fitzmaurice, M. Kelly, H. Ogunleye,J. Belter, and C. H. Song, A web-based electrical and electronics remotewiring and measurement laboratory (RwmLAB) instrument, IEEE Trans.Instrum. Meas., vol. 54, no. 1, pp. 3844, Feb. 2005.[35] D. Hercog, B. Gergic, S. Uran, and K. Jezernik, A DSP-based remotecontrol laboratory, IEEE Trans. Instrum. Meas., vol. 54, no. 6, pp. 30573068, Dec. 2007.[36] G. Donzellini and D. Ponta, A simulation environment for e-learning indigital design, IEEE Trans. Ind. Electron., vol. 54, no. 6, pp. 30783085,Dec. 2007.[37] C. Lazar and S. Carari, A remote-control engineering laboratory, IEEETrans. Ind. Electron., vol. 55, no. 6, pp. 23682375, Jun. 2008.[38] A. Mittal, C. Gupta, and A. Gupta, Addressing the bandwidth efficiency,control, and evaluation issues in software remote laboratory, IEEE Trans.Ind. Electron., vol. 55, no. 6, pp. 23262333, Jun. 2008.[39] M. J. Callaghan, J. Harkin, T. M. McGinnity, and L. P. Maguire, Intelli-gent user support in autonomous remote experimentation environments,IEEE Trans. Ind. Electron., vol. 55, no. 6, pp. 23552367, Jun. 2008.[40] S.-C. Wang and Y.-H. Liu, Software-reconfigurable e-learning platformfor power electronics courses, IEEE Trans. Ind. Electron., vol. 55, no. 6,pp. 24162424, Jun. 2008.[41] J. Ma and J. V. Nickerson, Hands-on, simulated, and remote laboratories:A comparative literature review, ACM Comput. Surv., vol. 38, no. 3,pp. 124, Sep. 2006.[42] E. Lehtinen, Computer supported collaborative learning: An approachto powerful learning environments, in Unraveling Basic Componentsand Dimensions of Powerful Learning Environments. Amsterdam, TheNetherlands: Elsevier, 2003, pp. 3553.Frederico Menine Schaf was born in Alegrete,Brazil, in 1979. He received the B.S. degree inelectrical engineering in 2002 from the UniversidadeFederal de Santa Maria, Santa Maria, Brazil, and theM.Sc. degree in electrical engineering in 2006, withemphasis on industrial automation and instrumen-tation, from the Federal University of Rio Grandedo Sul (UFRGS), Porto Alegre, Brazil, where he iscurrently working toward the Ph.D. degree.He is staying as a Guest Researcher with the Uni-versitt Bremen, Bremen, Germany. He developedseveral applications, tools, and didactic materials for Web-accessible experi-ments within the scope of the European Alfa II project called RExNet. His mainresearch interests include mixed reality, remote handling, multiagents systems,collaborative learning and work environments, Web and learning technologies,and modeling and simulation that are always focused on the field of engineeringeducation and training, where he has several publications.Carlos Eduardo Pereira was born in Porto Alegre,Brazil, in 1965. He received the B.S. degree in elec-trical engineering and the M.Sc. degree in computerscience from the Federal University of Rio Grandedo Sul (UFRGS), Porto Alegre, in 1987 and 1990,respectively, and the Dr.-Ing. degree in electrical en-gineering from the University of Stuttgart, Stuttgart,Germany, in 1995.He is currently an Associate Professor with theElectrical Engineering Department, UFRGS. His re-search focuses on methodologies and tool supportfor the development of distributed real-time embedded systems, with specialemphasis on industrial automation applications and the use of distributedobjects over industrial communication protocols.Dr. Pereira is the Chair of the International Federation of Automatic Control(IFAC) Technical Committee on Manufacturing Plant Control (TC 5.1). Hehas acted as a member of International Program Committees for severalconferences in the field of industrial automation, manufacturing, industrialprotocols, and real-time distributed object computing. He is also an AssociateEditor of the journal Control Engineering Practice. He is currently the Chair ofthe Brazilian Automation Society, the IFACs national member organization inBrazil. /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice

Recommended

View more >