Mobile Virtual Laboratory: Learning Digital Design - CiteSeerX

Mobile Virtual Laboratory: Learning Digital Design Vlado Glavini Faculty of Electrical Engineering and Computing University of Zagreb Unska 3, 10000 Zagreb, Croatia [email protected]

Mihael Kukec College of Applied Sciences Hallerova aleja 5, 42000 Varaždin, Croatia [email protected]

Abstract. An important part of learning digital design is the acquisition of knowledge through working experience in a laboratory environment, what will corroborate theoretical knowledge exposed elsewhere (e.g. in lectures, through written exercise solving etc.). This paper describes the foundation of a virtual laboratory available to every student at any moment and regardless of her/his whereabouts, following the concepts of universal access. We discuss the main issues in running a digital design lab, the essentials of contemporary mobile technology and the possible solutions to merge the two with the final objective being the development of a multiplatform system able to run both on (desktop) PCs and on mobile devices (terminals). We also discuss issues in the supporting HCI, especially in making the interaction as simple as possible. Keywords. adaptable user interfaces, mlearning, mobile HCI, universal access, virtual laboratory

1. Introduction Rapid development of new mobile technologies, communication-information infrastructure, and of widely-accepted e-learning systems has made it possible to introduce a new paradigm in the field of distance learning, which is denoted mobile learning or m-learning. Generally speaking, m-learning is the intersection of online learning and mobile computing, which promises the access to applications supporting learning anywhere and anytime [1], implementing the concepts of universal access [2]. The definition of mlearning is to learn knowledge ubiquitously and movably [3]. As with all other new paradigms m-learning has its own advantages and deficiencies. New and fast developing mobile computing devices

Sandi Ljubi Faculty of Engineering University of Rijeka Vukovarska 58, 51000 Rijeka, Croatia [email protected]

(mobile phones, pocket PCs, Tablet PCs, and PDAs) and mobile technology provide people with "wearable" computing ability to conveniently participate in distance learning environments [4]. Advantages of m-learning include mobility itself, better time-spending efficiency and more opportunities for peer collaboration. Furthermore, with the use of mlearning, distance education can encompass students with different levels of knowledge, capabilities and learning style categories, as well as anyone who requires mobile computer solutions that other devices cannot provide at the given moment or location. On the other hand, mobile learning disadvantages arise as a result of its general weaknesses, from psychological, pedagogical and technical viewpoints (as described in [5]). This especially applies to the most-frequently used mobile devices – mobile phones. General restrictive factors of these devices are small displays and their low resolutions, input limitations, limitations in accessing the Internet and the lack of standardization and compatibility. In order to make an effective use of an mlearning system, all available advantages and (we hope) temporary disadvantages are to be considered, along with the type and features of the application to be provided by m-learning means to the users. In this paper, we discuss a specific model of m-learning application, which is to support a mobile virtual laboratory for teaching digital design. The paper is structured as follows: Section 2 explains the motivation to develop such a laboratory, Section 3 shortly outlines available mobile devices and pertinent technologies, Section 4 delves into specific requirements for the mobile virtual laboratory model, and Section 5 discusses the relevant HCI issues.

325 th

Proceedings of the ITI 2007 29 Int. Conf. on Information Technology Interfaces, June 25-28, 2007, Cavtat, Croatia

the virtual laboratory, which we expect to become a very useful m-learning service.

2. Motivation The idea to build a mobile virtual laboratory comes as the result of several years of observation of students in several forms of laboratory training in digital design. Traditional training takes place in a specially equipped laboratory where students model and implement digital circuits using protoboards and integrated circuits, the correctness of their design being verified on standard measuring instruments, as shown in Fig. 1. Due to both a limited quantity of (relatively expensive) equipment and a usually large number of students, this form of laboratory training becomes increasingly ineffective and time-consuming.

Figure 1. Classic laboratory environment for teaching and training digital design

On the other hand, digital circuits modeling and design can be performed on (desktop) PCs by using appropriate development and simulation packages. Aside providing students with a new dimension in design, since it enforces the use of a suitable hardware description language (e.g. VHDL [6], see Fig. 2), the number of possible student assignments/projects can be increased, imagination being the only limit. Additionally, problems linked to failure of expensive measurement instruments are not present any more. However, let not be missed the problem of administration and coordination in such a computer laboratory, possibly involving quite a number of student groups. The issues referred to above influenced the modeling of a system that would enable distancebased laboratory training using either wireline interconnected computers (e-Lab), or wireless interconnected mobile devices (m-Lab). This combined functionality actually describes a complete model of a virtual laboratory (Fig. 3). This paper concentrates on the mobile aspect of

Figure 2. VHDL modeling and simulation of digital circuits

The idea of a mobile laboratory for learning and training digital design is supported by several factors, the first being the rapid growth of information and communication technology [7]. Mobile and handheld devices (PDAs and mobile phones of various types) are becoming both more powerful and versatile, having the ability to connect to the Internet and perform high data rate transfers. These characteristics, combined with decreasing prices, have made those devices very appealing and in fact quite common among the student population. On the other hand, the software running on these devices is predominantly designed to support the business world, with typical applications including time management, and communication and productivity tools [8]. Actually there is almost no software on mobile devices that would support any kind of learning or even curriculum delivery, although some of them run applications for reading e-books (which are mostly third-party applications). However, there already exist studies in m-learning (e.g. [9]), as well as working m-learning applications (e.g. [10]).

3. Mobile devices and available technologies This paper focuses on mobile terminals like PDAs and mobile phones, and their combination – smartphones. All three categories represent small, lightweight and portable devices, which are omnipresent, and constantly with their owners, what especially is the case with mobile phones. This makes them ideal platforms for supporting m-learning applications, so that the users are able to initiate and run specific applications at any moment of their free time.

326

From the application development perspective mobile phones can generally be divided into two groups, the important difference between them being their ability to support application development (see Fig. 4). The first group consists of devices controlled by proprietary operating systems (OSs), which are usually model specific (e.g. Nokia mobile phones running on the "Series 40" platform [11]), while the other of smart-/mobile phones (where a distinct line between these two device types is somewhat blurred), which are usually controlled by more complex OSs (e.g. Symbian, Windows Mobile, Palm OS, or Blackberry OS).

solving this seems to be the use of environments as the Java 2 Micro Edition (J2ME) platform [12], which is designed for development of applications for small devices of limited capabilities ("small" and "limited" with respect to usual resources of a standard PC), like mobile phones, smartphones and PDAs. Thus far almost every mobile phone is Java-enabled, and is therefore capable of executing a J2ME application. J2ME is platform-independent and makes the creation of mobile, Web-based applications possible [13]. Proprietary OS (no API issued)

Smartphones

Mobile phones

PDAs

Symbian, Windows Mobile, Palm OS, Blackberry OS,… («open» API)

Figure 4. Handheld devices

Figure 3. Virtual laboratory model

The first group cannot accept non-proprietary applications because they represent technically "closed" systems, characterized by publicly inaccessible OS information and undisclosed APIs. Applications for a specific OS are "hardwired" and usually consist of some simple programs like calendars, clocks or appointment lists. The other group relies on open APIs, hence enables users to install non-proprietary applications on the respective mobile devices. Additionally, manufacturers of these latter devices even encourage development of applications for their OSs, supporting widely familiar programming languages like C++, Java and some others. The above discussion illustrates the problem of diversity of both OSs and associated program tools for application development, which makes very difficult the creation of a general, multiplatform application to be run on an arbitrary mobile phone. The problem is accentuated if only the manufacturers are able and authorized to create applications for certain devices. For the moment being, the best way of

PDA devices generally have more powerful processors and larger memories hence no problem arises with applications to be shared with mobile phones. J2ME support is implemented in the most common PDA operating systems: PalmOS and Windows Mobile. PalmOS supports J2ME through IBM WebSphere Everyplace Micro Environment Java J2ME runtime, while most of the current Windows Mobile based PDAs have third-party Java Virtual Machines pre-installed [14]. There seems to be quite a homogeneous situation regarding J2ME platform support in mobile phones since it usually is already installed in the OS. This is however not the case with PDAs where sometimes the user has to install the support on his own and in some cases pay extra for it. Such a situation is going to be improved with packages like phoneME – an open-source project trying to further expand the usage of Java™ Platform, Micro Edition (Java ME platform) technology in the mobile handset market [15]. MIDP (1.0) CLDC (1.0)

Figure 5. Sample J2ME Stack

When speaking of J2ME support, we assume that the J2ME software stack, see Fig. 5, is implemented in the mobile device. The lower

327

layer CLDC (Connected, Limited Device Configuration) defines a configuration which determines all that is needed in order to have basic Java Technology support in a compliant device, while the higher one MIDP (Mobile Information Device Profile) defines the profiles which provide industry-specific or specialized APIs on top of the underlying configuration. CLDC and MIDP together specify the core VM and Java language features plus the wireless and mobile-specific APIs required for mobile devices [16]. Another advantage of the J2ME platform is the simplicity in implementing the model of interactive communication between the mobile device and the distant Internet-connected server. Since all MIDP implementations must have support for HTTP, interactive exchange of data between these two entities can be created as a cooperation of MIDlet and Servlet applications, as shown in Fig. 6. Here MIDlet represents any MIDP application which can be realized on a mobile device while Servlet represents a server application based on the widely-accepted J2SE platform.

deliver interactive contents for learning and training, (ii) to guide through the entire learning process (including an evaluation mechanism), and (iii) to administer all users. The delivered learning contents usually consists of multimedia materials which combine texts, pictures and audio files. More complex interactive applications enable more sophisticated tasks, like laboratory training or testing. Delivered content could be shaped as a set of Mobile Interactive Learning Objects (MILO), see [17], of a high abstraction level, hence having a wide area of possible implementation. To work in a specific field it is necessary, and possible, to develop interactive objects that would support the studied materials. A MILO set used in teaching and learning digital design, along with the associated laboratory training, can be further expanded with a tool for modeling and simulating digital circuitry. Such tool should enable the users not only to model and simulate, but also to group simple logic models into more complex ones, with the extra feature of storing them for future usage. Furthermore, the system must make it possible for the author of the objects to create certain sets of assignments and exercises that can further be used to automatically generate assignment instances later on. It should also enable the author to implement an automatic evaluating system for solutions received. According to evaluation results the next action can be set: if the solution is not satisfactory, the system can inform the user and with an adequate help, meaning extra information, gradually guide her/him to the correct solution of the exercise. If on the other hand the solution is completely, or partially, correct, it can further examine the user or it can direct her/him to learn other parts of the curriculum and do other laboratory assignments.

Figure 6. MIDlet – Servlet communication model

4. Requirements for mobile virtual laboratory model In the remainder of the paper we explore the virtual laboratory model by discussing the solutions available for obtaining a multiplatform system with applications able to run on PCs as well as mobile devices. System tasks are (i) to

328

Figure 7. Learning/training cycle in a mobile laboratory environment

The learning and training cycle for a mobile virtual laboratory is shown in Fig. 7. It illustrates a model of a laboratory system where the support for simulating logic circuitry being modeled on the user's side, is on the mobile device itself. This means that in such a model implementation that particular component must be separately developed, as e.g. a lightweight Java simulator whose executive code must not exceed the memory limits of target devices. As opposed to such a model it is possible to use an already complete solution instead of creating one's own; this solution is regularly too demanding for the memory of handheld devices, and in most cases is not realized with Java technology, hence it is transferred to the server side of the virtual laboratory system, see Fig. 8. The user of the mobile device receives the object describing her/his laboratory assignment, goes through circuitry modeling using the application running on her/his device and eventually forwards the developed model onto the m-service server. The server based simulator is capable of interpreting the received data and returns the generated results to the user application, which in turn visualizes them. In the process, the user also receives information on possible modeling errors. Upon a user's mistake the system can decide whether to re-examine her/him, set easier laboratory assignments, or "arrange" for an additional delivery of learning materials concerning the specific laboratory exercise.

Figure 8. Laboratory task within virtual laboratory model with server-side simulator

The assignments given, and especially laboratory exercises, must make use of interactive capabilities of the device, which usually involves a mobile game-based learning approach [18], where special emphasis is given on the system having all the appeal of a game but

at the same time retaining its primary learning function.

5. HCI issues Using a learning system in a mobile environment is not effective if obstacles are not overcome in the interaction with the targeted user(s), which results from limited both input and output capabilities of mobile devices. For that purpose all mobile services, including m-learning ones, demand special attention being paid to interaction with the user [19]. Developing allpresent learning applications also involves developing an adjustable interface to work with different devices where the application is used [20].

5.1. Device-based adaptable user interface The final objective being the development of a virtual mobile laboratory to be accessed from a (desktop) PC, a mobile phone and a PDA, it should be noted that all the mentioned devices have their own specific features which are significantly different, what brings us to the issue of creating a unique application having functionally the same or very similar user interface (UI). Section 3 discusses why Java (J2SE and J2ME) is the most suitable technology and the logical choice for development of such an application. There are however technologies enabling development of applications with a unified UI for different types of handheld devices [21, 22], but they only offer APIs for supporting general-purpose widgets exemplified by graphical UI (GUI) elements such as buttons or editable fields. This is not sufficient in our case since the user interface must show more complex graphic elements like symbols of digital circuits and virtual devices. Besides, these technologies are not entirely useful because they are not compatible with the Java technology although they were made to resemble it. A mobile virtual laboratory application can be divided into three functional blocks: (i) user interface and interaction with the user, (ii) communication between interface and simulation components, and (iii) components for simulation of digital circuitry. As a detailed analysis of communication and simulation components is beyond the scope of this paper, in the following we will concentrate on the user interface and the associated interaction.

329

In order to present the application interface in a device-independent way, it is necessary to develop a virtualizing layer between the device and the system core, its primary function being to receive user commands across the interface. While doing that, the virtualizing layer must be able to interpret actions originating from all three device categories and relay them into the system core in a unified way. This is because the same action is performed differently on different types of devices, using different interaction styles. E.g. a desktop user will mostly use the mouse and to a lesser extent the keyboard, the user of a mobile phone will use numeric and navigation keys or even micro-joysticks, whereas a PDA user will perform the same actions by using a "soft" keyboard on a touch-sensitive display. The second function of the virtualizing layer is the presentation of the interface on a specific device display. The system core identifies the user's actions and, in accordance with their meaning, generates the interface on the canvas of the virtualizing layer, hence enabling it to create the image on the target display. Fig. 9 shows the implementation details of the system core and the virtualizing layer. DigitalLogicSimulator

UserInterfaceEngine

AppletCanvas +Graphics

+DrawOnVirtualCanvas() +keyPressed() +pointerPressed() +mouseMoved()

+keyPressed() +mouseMoved()

AppletVirtualCanvas

+keyPressed() +pointerMoved()

MIDLetVirtualCanvas

-Graphics +DrawAppletUI() +setGraphics()

MIDLetCanvas +Graphics

-Graphics

Virtualising layer

+DrawMidletUI() +setGraphics()

VirtualCanvas -Graphics +DrawUI() +setGraphics()

Figure 9. UML class diagram for adaptable user interface

It can be seen that the system core is represented as a UserInterfaceEngine class, whereas the virtualizing layer is represented by the cooperation of this and the VirtualCanvas class. AppletCanvas and MIDLetCanvas represent classes which describe a computer monitor and a mobile phone display,

respectively. System operation follows the next steps: a PC application (e.g. an Applet), and a mobile phone/PDA MIDLet will implement no interface functionality (neither in their own classes nor in any directly inherited classes) except the communication with the core, i.e. the UserInterfaceEngine object. This latter object processes users' actions, like pressing a key or moving the mouse, and in accordance with the new situation generates the interface in a uniform way disregarding the type of the access device. The VirtualCanvas object generates the image on the specific access device display. Hence, the virtualizing layer is realized through the cooperation of the core, the UserInterfaceEngine, and the VirtualCanvas objects, which reconciles the differences while generating the image on different devices.

5.2. Towards an intelligent interface The ultimate objective to be addressed in mlearning is clearly UI automatic adaptation in accordance with both user preferences and level of knowledge. In order to make system use as simple and efficient as possible, the user interface should actively monitor the user while interacting with the system. This involves installing suitable adaptation algorithms which should operate "in the background" and, based on data thus collected, change the momentary interface look. Applicable algorithms can be (i) recency-based, monitoring the most-frequently used commands during the entire interaction, or (ii) frequencybased, calculating the frequency of usage of certain commands in shorter interactive cycles. E.g., [23] shows how it is possible to realize different types of interface adaptation basing on these algorithms. Furthermore, user interface adaptation need not develop only towards personalization of interaction but also towards generating very simple – minimalistic – user interfaces. Automatic adaptation already touches upon some intelligent procedures in the system-user interactivity domain by modeling a user behavior profile, cf. [24]. This profile, which describes the preferred way of interaction, can also be used in anticipating future user objectives in specific tasks of her/his work process. Such objectives anticipation allows greater interaction efficiency and speed by introducing system interface "shortcuts", as well as auxiliary information on the interface associated to the current task. Along

330

with behavior profile modeling, it is also possible to model a user knowledge profile (cf. [25]), containing and updating user information which would help the system to determine the category of her/his both knowledge and skills in certain fields of expertise (e.g. beginner, intermediate, advanced, expert). Based on these information it would be possible for the user interface to make the work easier for each user by automatically adjusting itself during the usage of the system.

adaptation in agreement with the type of the access device. We are continuing our work on the implementation of the model presented by primarily focusing on the development of mobile device applications, where we expect to fully show the advantages of the described mobile HCI. These applications will be tested in various working environments, with the results thus accomplished being the measure of user efficiency and satisfaction analysis.

6. Conclusion and future work 7. Acknowledgements M-learning systems will replace neither the traditional way of learning nor the already widely-accepted e-learning services. They will however certainly become an additional benefit in the comprehensive process of lifetime learning. In the paper we considered the possibility of performing simple laboratory exercises using mobile devices, what also brings us closer to implementing a full-scale mobile virtual laboratory. Capabilities of mobile devices, their input and output, are still not at a level where their applications could compete with some standard desktop applications. However, the contemporary development of mobile technologies and the associated informationcommunication infrastructure do make possible the introduction of m-learning through delivering various multimedia and interactive contents for learning and interaction between mobile users and m-learning servers. Developing a virtual laboratory system which would offer on the one hand e-learning services, usually on desktop computers, but also on the other adequate mlearning services for users of handheld mobile devices, can definitely be considered a significant breakthrough in the field of computer supported education. This paper proposes a model of such a virtual laboratory, which supports m-learning and training in the field of digital design by (i) delivering interactive multimedia learning contents to mobile devices, (ii) modeling logic circuits with an application on the mobile device, (iii) simulating circuit models on a distant mlearning server or on the mobile device itself, and (iv) visualizing simulation results in accordance with the capabilities of the used mobile device. HCI aspects of the system are also discussed, addressing long-term objectives as well as a possible solution for user interface

This paper describes the results of research being carried out within the project 0360361994-1995 Universal Middleware Platform for e-Learning Systems funded by the Ministry of Science, Education and Sports of the Republic of Croatia.

8. References [1] Tatar D, Roschelle J, Vahey P. Handhelds Go to School: Lessons Learned. IEEE Computer 2003; 36(9): 3-37. [2] Stephanidis C. Editorial. Int'l J. Universal Access in the Information Society 2001; 1(1): 1-3. [3] Ogata H, Yano Y. Knowledge Awareness Map for Computer-Supported Ubiquitous Language-learning. Proc. 2nd IEEE Int'l Workshop Wireless and Mobile Technologies in Education (WMTE’04); 2004 Mar 23-25; Taoyuan, Taiwan. IEEE Computer Society; 2004. pp. 19-26. [4] Mikic F, Anido L. Towards a Standard for Mobile E-Learning. Proc. Int'l Conf. Networking, Int'l Conf. Systems and Int'l Conf. Mobile Communications and Learning Technologies (ICN/ICONS/MCL'06); 2006 Apr 27-29; Mauritius. IEEE Computer Society; 2006. pp. 217. [5] Shudong W, Higgins M. Limitations of Mobile Phone Learning. Proc. IEEE Int'l Workshop Wireless and Mobile Technologies in Education (WMTE'05); 2005 Nov 28-30; Tokushima, Japan. IEEE Computer Society; 2005. pp. 179-181. [6] Brown S, Vranesic Z. Fundamentals of Digital Logic with VHDL Design. Boston, MA: McGraw-Hill; 2000. [7] Georgiev T, Georgieva E, Smrikarov A. MLearning - a New Stage of -Learning. Proc. 5th Int'l Conf. Computer Systems and

331

Technologies (CompSysTech’04); 2004 Jun 17-18; Rousse, Bulgaria. ACM Press; 2004. pp. 1-5. [8] Corlett D, Sharples M, Chan T, Bull S. A Mobile Learning Organiser for University Students. Proc. 2nd IEEE Int'l Workshop Wireless and Mobile Technologies in Education (WMTE'04); 2004 Mar 23-25; Taoyuan, Taiwan. IEEE Computer Society; 2004. pp. 35-42. [9] Bull S, Cui Y, Röbig H, Sharples M. Adapting to Different Needs in Different Locations: Handheld Computers in University Education. Proc. IEEE Int'l Workshop Wireless and Mobile Technologies in Education (WMTE'05); 2005 Nov 28-30; Tokushima, Japan. IEEE Computer Society; 2005. pp. 48-52. [10] Holzinger A, Nischelwitzer A, Meisenberger M. The Mobile Learning Engine (MLE) – A Mobile, Computer-aided, Multimedia-based Learning Application. Proc. Multimedia Applications in Education Conference (MApEC'04); 2004 Sep 13-15; Graz, Austria. pp. 52-58. [11] Nokia series 40 developer platform. http://forum.nokia.com/main/platforms/s40/ [02/22/2007] [12] Shaylor N, Simon DN, Bush WR. A Java Virtual Machine Architecture for Very Small Devices. Proc. ACM SIGPLAN Conf. Language, Compiler, and Tool for Embedded Systems (LCTES'03); 2003 Jun 11-13; San Diego, California, USA; ACM Press; 2003. pp. 34-41. [13] Holzinger A, Nischelwitzer A, Meisenberger M. Lifelong-learning Support by m-learning: Example Scenarios. ACM eLearn Magazine 2005; 11: 2. [14] A Guide to Windows Mobile Programming for Symbian OS Developers. http://msdn2.microsoft.com/en_us/library/ aa454908.aspx [02/22/2007] [15] Java ME phone platform development. https://phoneme.dev.java.net [02/22/2007] [16] Day B. Developing Wireless Applications using the Java 2 Platform, Micro Edition. Sun Microsystems, Inc.; 2001.

http://developers.sun.com/techtopics/mobilit y/getstart/articles/wirelessdev [02/22/2007] [17] Holzinger A, Nischelwitzer A, Meisenberger M. Mobile Phones as a Challenge for mlearning: Examples for Mobile Interactive Learning Objects (MILOs). Proc. Third IEEE Int'l Conf. Pervasive Computing and Communications (PerCom'05); 2005 Mar 812; Kauai Island, Hawaii, USA. IEEE Computer Society; 2005. pp. 307-311. [18] Cogoi C, Sangiorgi D, Shahin K. mGBL – mobile Game-Based Learning: Perspectives and Usage in Learning and Career Guidance Topics. eLearning Papers 2006; vol 1. http://www.elearningpapers.eu [02/22/2007] [19] Chittaro L. HCI Aspects of Mobile Devices and Services. Personal Ubiquitous Computing Magazine 2004; 8: 69-70. [20] Repo P. Facilitating User Interface Adaptation to Mobile Devices. Proc. Third Nordic Conf. Human-Computer Interaction (NordiCHI'04); 2004 Oct 23-27; Tampere, Finland. ACM Press; 2004. pp. 433-436. [21] Waba programming platform. http://www.wabasoft.com [02/22/2007] [22] SuperWaba. http://www.superwaba.com.br [02/22/2007] [23] Gajos KZ, Czerwinski M, Tan DS, Weld DS. Exploring the Design Space for Adaptive Graphical User Interfaces. Proc. Working Conf. Advanced Visual Interfaces (AVI'06); 2006 May 23-26; Venezia, Italy. ACM Press; 2006. pp. 201-208. [24] Grani A, Glavini V. Adaptive Interface of a Human Teacher Emulator. Proc. 11th Int'l Conf. Human-Computer Interaction (HCII 2005). 3rd Int'l Conf. Universal Access in Human-Computer Interaction: Exploring New Dimensions (UAHCI 2005); 2005 Jul 22-27; Las Vegas, NV, USA. Salvendy, G. Ed.; 2005. pp. 1-10. [25] Rosi M, Stankov S, Glavini V. Personal Agents in Distance Learning Systems. In: Elmenreich W, Tenreiro Machado JA, Rudas IJ, editors. Intelligent Systems at the Service of Mankind, Vol. I. Augsburg: Ubooks, 2004.

332