Design of a New Mobile-Optimized Remote Laboratory ... - IEEE Xplore

2382

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 64, NO. 3, MARCH 2017

Design of a New Mobile-Optimized Remote Laboratory Application Architecture for M-Learning Ning Wang, Student Member, IEEE, Xuemin Chen, Senior Member, IEEE, Gangbing Song, Member, IEEE, Qianlong Lan, and Hamid R. Parsaei

Abstract—As mobile learning (M-Learning) has demonstrated increasing impacts on online education, more and more mobile applications are designed and developed for the M-Learning. In this paper, a new mobile-optimized application architecture using Ionic framework is proposed to integrate the remote laboratory into mobile environment for the M-Learning. With this mobile-optimized application architecture, remote experiment applications can use a common codebase to deploy native-like applications on many different mobile platforms such as iOS, Android, Windows Mobile, and Blackberry. To demonstrate the effectiveness of the proposed new architecture for M-Learning, an innovative remote networked proportional–integral–derivative control experiment has been successfully implemented based on this new application architecture. The performance is validated by the Baidu mobile cloud testing bed. Index Terms—Ionic framework, mobile learning (Mlearning), mobile-optimized application architecture, remote laboratory, unified framework.

I. INTRODUCTION URING recent decades, the rapid development of communications and wireless technologies have resulted in mobile devices (e.g., smartphones and tablets) becoming widely available, more convenient, and less expensive [1]. Mobile learning (M-Learning) [2], which is the delivery of learning, education, or learning support on mobile phones or tablets, has played an important role in E-Learning. According to research by Global Industry Analysts published in 2014 [3], it states that

D

Manuscript received April 15, 2016; revised July 7, 2016 and August 31, 2016; accepted September 24, 2016. Date of publication October 21, 2016; date of current version February 9, 2017. This work was supported by the Qatar National Research Fund under Grant NPRP 4-892-2-335. N. Wang is with the Department of Electrical and Computer Engineering, University of Houston, Houston, TX 77204 USA (e-mail: [email protected]). X. Chen is with the Department of Engineering, Texas Southern University, Houston, TX 77004 USA (e-mail: [email protected]). G. Song is with the Department of Mechanical Engineering, University of Houston, Houston, TX 77204 USA (e-mail: [email protected]). Q. Lan is with the Department of Computer Science, Texas Southern University, Houston, TX 77004 USA (e-mail: q.lan7324@ student.tsu.edu). H. R. Parsaei is with the Department of Mechanical Engineering, Texas A&M University at Qatar, Doha, Qatar (e-mail: hamid.parsaei@qatar. tamu.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIE.2016.2620102

“the E-Learning market is one of the most rapidly growing sectors in the global education industry.” E-Learning, which was instituted on desktops in the beginning, has gradually shifted its base to portable tablets and smartphones [4]. Thus, more and more learning approaches and learning systems are configured and integrated for M-Learning [5]. Examples include a microlecture M-Learning system at Guangdong University of Technology [6] and a smart learning mobile system for collaborative M-learning at BaekSeok Culture University [7] among others. It is expected that the teaching and learning actives will move more and more outside the classroom into the learners’ personal environment both real and virtual mediated by mobile devices [8]. Thus, it continues to be a major technology trend as we move in future. Because of its effectiveness, flexibility, and cost saving, the remote laboratory technology, as an important component of online learning, has made great progress. Many applications based on remote laboratory technology are being recognized in science, technology, engineering, and math education [9]. Till now, some of the most prominent examples of remote laboratories that have been successfully implemented include the MIT iLab [10], the WebLab-Deusto [11], the Networked Control System Laboratory (NCSLab) [12], the improved NCSLab 3-D [13], and the eComLab [14]. Based on these successful examples, it has proven that remote laboratories can be highly effective tools in helping a wider range of students, regardless of geographical restrictions, to obtain practical experience needed for competency in science and engineering. The location independent access of a remote laboratory is especially useful in scenarios where space is limited, or for distance education [15]. To offer students a more flexible way to access remote laboratory, instead of forcing the learners to sit in front of a fixed computer to use a location-independent environment for experimentation, a technology that is suitable for presenting remote laboratory on mobile devices becomes essential. In addition, integrating the remote laboratory into the mobile devices can offer students more flexible approach to learning and produce better outcomes as pointed out by May et al. [16] and Silva et al. [17]. So far, most of current research interests in M-Learning mainly focused on the various learning theories, there are only a smaller number of research works that focused on the design of framework and the mobile devices technologies that are

0278-0046 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

WANG et al.: DESIGN OF A NEW MOBILE-OPTIMIZED REMOTE LABORATORY APPLICATION ARCHITECTURE FOR M-LEARNING

2383

TABLE I EXAMPLES OF MOBILE REMOTE LABORATORY APPLICATION Mobile remote laboratory example

Institution

Year

Mobile application type

Description from vendor

1. An Android-based remote laboratory application [24] 2. A field fluorometer experiment using smartphone [25] 3. J2ME-based mobile laboratory [26]

MIT

2012

Native mobile application

University of Sydney

2015


Princess Sumaya University

2008


4. Virtual and remote laboratory [27]

National University of Distance Education (UNED)

2013


5. OMF measurement laboratory [28]

University of Malaga

2016


6. 3-D mobile application for remote experiments [29]

Indiana University Northwest

2016


Develop a remote Android client application based on MIT iLab shared architecture. Develop a novel portable fluorometer using android-based application framework. Develop a mobile remote lab application based on J2ME framework. Develop a remote lab application to support portable devices based on EJS (Easy Java Simulation). Add a measurement and monitoring tool TestelDroid into android devices to control remote experiments. Develop a 3-D remote laboratory experiments application based on 3-D mobile Augmented Reality Interface technology. Develop a remote laboratory using AJAX technology based on WebLab-Deusto architecture. Develop a remote laboratory using HTML 5 technology and based on WebLab-Deusto platform. Develop a remote laboratory using HTML 5 technology and based on Moodle LMS. Develop mobile remote experiment application using HTML5 and jQuery Mobile framework. Develop a mobile remote lab application using HTML 5 technology and Node.js server. Using “iSES Remote Lab SDK” for Arduino-UNO and web technology to develop a mobile remote laboratory application.

7. A web-based mobile application based University of Deusto on WebLab-Deusto architecture [30] 8. A Mobile-accessed remote laboratory Blekinge Institute of Technology (BTH) application for VISIR [31] 9. PeTEX platform [32] TU Dortmund University

2008

Web mobile application

2006


2011

Web Mobile Application

10. RExLab [33] 11. Remote Web-based mobile control laboratory [34] 12.iSES REMOTE Lab SDK for Smartphone [35]

Federal University of Santa Catarina (UFSC) Complutense University

2016


2015


University of Prague

2016


compatible for M-learning systems [18], [19]. How to design and implement a mobile-optimized and easy-to-use application for M-learning has become an emerging research topic [20], [21]. Generally, two approaches, the native mobile application and the web-based application, are used to integrate the remote laboratory into mobile devices. 1) Native remote laboratory application for mobile devices: Native remote laboratory applications are developed using different native codes, different tools, build systems, application programming interfaces (APIs), and mobile devices with different capabilities for different platforms, such as Apple iOS, Android, Windows Mobile. Meanwhile, native remote laboratory applications are compiled and they directly call the underlying APIs of the different platforms. Although the native remote laboratory applications can achieve best performance on mobile devices with different platforms, it is hard to implement the cross-platform interface [22]. In Table I, examples 1–6 are the native mobile remote laboratory applications or mobile industrial applications that were developed with native application approach for different mobile systems. However, they are hard to be ported to other mobile systems [23]. 2) Web-based remote laboratory application for mobile devices: Web-based remote laboratory applications normally are created in HyperText Markup Language (HTML), Cascading Style Sheets (CSS), and JavaScript; and they run in the different web browsers (such as Safari, Chrome, Microsoft IE, Firefox.) [36]. In Table I, the examples 7–12 are the web-based mobile remote laboratory applications. Example 8 used jQuery mobile framework to implement the mobile remote laboratory application. Examples 9–12 were implemented by

web development technology as well. All of these mobile remote laboratory applications can be easily run on different mobile phone systems via different web browsers; but their performance depends on JavaScript rendering and mobile web browsers. With the rapid development of web technology, the web application running performance has been significantly improved. However, for some special user interfaces (UI), e.g., example 6 in Table I, the web application running performance is still a major issue. In addition, some mobile applications need to use sensors in mobile devices, e.g., examples 2 and 5 in Table I. Thus, native application is the only option. The other main drawback of web application is the limited access to mobile device hardware. Therefore, how to design a mobile remote laboratory application architecture, which has the native-like performance and native functional capability, also can run on cross platform easily and take low-cost and development efforts like web applications, is already an essential issue [22], [23], [36]. To address this essential issue for better support the M-Learning, a new mobileoptimized remote laboratory application architecture based on Ionic framework is proposed and implemented in this paper. To the authors’ best knowledge, that is the first study to build a mobile-optimized application architecture for remote laboratory application development based on Ionic framework. To demonstrate the feasibility of this new architecture, an innovative remote-networked proportional–integral–derivative (PID) control experiment based on such an architecture is designed and implemented. The Baidu Mobile cloud testing tool is used to test the performance of the new architecture. Since smartphone has been widely used in industry such as a field fluorometer experiment using smartphone [25], OMF measurement remote experiment using android [28], the new mobile-optimized

2384


remote laboratory application architecture can provide a new development tool for mobile industrial electronic devices with less efforts. The rest of this paper is organized as follows. In Section II, previous work on the unified framework for remote laboratory development and the comparison of mobile development frameworks is introduced. To upgrade the unified framework, a new methodology for optimized mobile applications based on Ionic framework is presented in Section III. In Section IV, a new mobile-optimized remote PID motor speed control experiment application for M-Learning is presented. The future works are summarized in Section V. Concluding remarks are drawn in Section VI. II. PREVIOUS WORKS A unified framework proposed in [37]–[39] has solved several critical issues to improve the remote experiment performance and user experience. The remote laboratory based on the unified framework can provide the students real-time video and realtime data transmission without software plugins and the firewall issue. To integrate the advantages of unified framework into the mobile environment for better supporting M-Learning, an efficient and stable mobile framework must be selected as the foundational framework. To select the efficient and stable mobile framework for remote experiment, the comparison of mobile development frameworks is also introduced in this section. A. Unified Framework for Remote Laboratory Development For the remote laboratory implementation, many of the solutions used Web 2.0 and web services/service-oriented architecture (SOA) technology to produce the better quality remote laboratories without software plugins [40]. However, the main drawback of web services is slow data transmission resulting in poor support for real-time data communication [40], [41]. To improve the real-time data transmission and keep the advantages of web services, a unified framework for remote laboratory development was proposed and implemented. This unified framework provides the UI level and platform level APIs to integrate and implement the remote experiments. The unified framework is based on the combination solution of both Apache web engine and Node.js web engine, and it uses a Node-HTTP-based technology to resolve the real-time communication between experiment equipment with end users [37]. The subsequent iteration of the design resolved challenges of developing cross-browser and cross-device web UI as an improvement to the unified framework [38], [39]. The system architecture of the unified framework is shown in Fig. 1. This unified framework is based on the Web 2.0 technology and includes three parts: client web application, server application, and experiment control application. 1) Client Web Application: The client web application runs on most of current popular browsers, and is based on HTML, CSS, and JQuery/JQuery-Mobile JavaScript libraries. Further, it uses the server-based Mashup technology for UI development.

Fig. 1.

Architecture of the unified framework [37].

2) Server Application: To improve the web services technology performance issue, for the server application implementation, the combination solution of both Apache HTTP web engine and Node.js web engine implements the real-time communication between the experiment hardware and end users. Apache HTTP web engine is the world’s most used web server software and supports a variety of UI features development [42]. Node.js is an open source, cross-platform runtime environment for developing server-side web applications. It also enables web developers to create an entire web application in JavaScript for both server side and browser side. In the Node.js server-side software system, Socket.IO, a JavaScript library for real-time web applications, is used to support real-time communication between server side and client side [43], [44]. Thus, the server application is directly built on the top of an Apache HTTP web engine, a Node.js web engine, and a MySQL database. For the database operations, only the Apache server communicates with MySQL using PHP and SQL for user management and scheduling system. The experiment data are directly saved into the file system using Node.js. Meanwhile, the operation system of the server uses CentOS for better support the server application. 3) Experiment Control Application: The experiment control application is developed with the Laboratory Virtual Instrument Engineering Workbench (LabVIEW), and runs on a workstation with Windows OS. To implement real-time communication between the client application and the LabVIEW experiment equipment control application, a new Socket.IObased application transmission protocol, LtoN (LabVIEW to Node.js), was designed and developed in the unified framework [38]. This new protocol is designed and implemented based on Socket.IO protocol. Socket.IO is the module of Node.js, and is designed based on web socket protocol. For the entire framework, three vital technologies were used. These technologies include: 1) Socket.IO protocol and Node-HTTP-Proxy used for experiment data and control commands transmission and traversing firewall [38];


TABLE II COMPARISON BETWEEN DIFFERENT MOBILE APPLICATIONS Features Running performance User interface Development time Development cost Native-like feature Cross-platform porting Average

Optimized applications Native applications Web applications 4 4 3 3 4 5 3.8

5 5 2 2 5 1 3.3

3 4 5 5 1 5 3.8

2) Novel video transmission approach is based on HTTP live streaming (HLS) protocol for real-time system monitoring [39]; and 3) Server-based Mashup technology is used for UI implementation. This unified framework has been used to implement several remote experiments for engineering education. For example, the new smart vibration platform (SVP) remote experiment is now used to teach students in mechanical engineering (ME) courses as reported in [45]. The remote SVP experiment offered students hands-on experience on structural vibration control by using a magnetorheological and shape memory alloy (SMA) braces to control the vibration of a one story model. B. Mobile Application Development Framework Selection Currently, most of the mobile remote laboratory applications are implemented using native application approach and webbased application approach. With the continuous improvement of mobile application development technology, developers are migrating to mobile-optimized application development tools such as, PhoneGap, jQuery Mobile, Adobe Air, Titanium, to reduce the cost of development and reach out to maximum users across several platforms [46]. Mobile-optimized application is a mix of native and web technologies that are leveraged to deliver a mix of web content and native capabilities. Native mobile applications are developed for one platform and can take full advantage of mobile device capabilities. Web-based applications are not exactly mobile applications but are websites that are mobile formalized. However, native features such as the interaction with device sensors cannot be ignored. Mobile-optimized application can be most suited for cross-platform requirements, since the same HTML content needs to be accessed from different mobile platforms. As compared with web applications, the mobile-optimized application has better performance that supports cross-platform development, and has similar native functional capability with native application. Moreover, the code portability of mobile-optimized application is better than native applications. Meanwhile, the development cost of mobileoptimized application is lower than native application. The comparison between different applications is shown in Table II. In Table II, the evaluation values are presented on a scale of one (very poor) to five (excellent) [47]. With the HTML5 technology development, more and more different mobile-optimized application frameworks for developing and building mobile applications emerge endlessly.

2385

An overview of the different cross-platform tools is shown in the Fig. 2. From the result of our literature review, the PhoneGap framework is the best framework for mobile-optimized application development based on [48]–[51]. The PhoneGap is an open source framework that provides developers with an environment where developers can create applications in HTML, CSS, and JavaScript and still call native device features and sensors via a common JavaScript API. The PhoneGap framework contains the native code pieces to interact with the underlying operating system and pass information back to the JavaScript app running in the WebView container. However, to deliver better user experience, Ionic framework emerges as the first full-stack service platform based on the PhoneGap framework to build and scale the mobile applications with HTML5 technology, and Ionic framework can support most of current popular mobile systems, e.g., Android, iOS, Windows Mobile, Blackberry, Amazon Fire OS, Firefox OS, Ubuntu Mobile OS, and Tizen [52]. Ionic framework is mainly focused on the end users looking, feeling, and user experience interaction of the mobile applications [53]. The Ionic framework is not a replacement of the PhoneGap framework. Instead, Ionic framework simply fits in well with these projects that need to simplify the complex UI part of the mobile applications. Meanwhile, it also offers a library of mobile-optimized HTML, CSS, and JavaScript components, gestures, and tools to build highly interactive applications. The Ionic framework works are based on two kernel components, Apache Cordova, and AngularJS. From Table III, the result of comparison with other frameworks shows that the Ionic framework has the unique advantage. Ionic framework enables high native-like performance, around 70%. Meanwhile, it can be supported by both of the two development environments, Apache Cordova framework and AngularJS framework. Based on Apache Cordova and AngularJS, the mobile-optimized applications, which are developed by JavaScript language, can be built without any native code (e.g., Java, Objective-C) in mobile devices [48]. In this way, Ionic framework can be used to implement interactive mobile applications with the unified framework smoothly on different mobile operation systems. Therefore, the Ionic framework can be the best candidate for the foundational platform of mobile-optimized application architecture. III. NEW MOBILE-OPTIMIZED APPLICATION ARCHITECTURE To answer our research question, “How to develop a mobile remote laboratory application running cross platform easily with competitive running performance and hardware accessibility as native application?”, a new mobile-optimized application architecture is proposed. This new architecture, which combines Ionic framework with the unified framework together, can have the advantages of both two frameworks. We expect it has better support for M-Learning. A. Propose a New Mobile-Optimized Application Architecture The mobile-optimized application architecture includes two layers, optimized application layer and unified framework layer as shown in Fig. 3.

2386

Fig. 2.


Usage distribution among cross-platform tool users in 2013, according to market analysis by Research2guidance [51].

TABLE III COMPARISON BETWEEN DIFFERENT FRAMEWORKS Framework 1. Ionic framework 2. Onsen UI 3. Framework 7 4. React native 5. jQuery mobile 6. Native script

Fig. 3.

Native-Like performance

Support Cordova/AngularJS

7/10 6/10 8/10 8/10 3/10 8/10

Yes/Yes No/Yes No/No No/No No/No No/No

Mobile-optimized remote laboratory architecture.

1) Mobile-Optimized Application Layer: The mobileoptimized application layer can support applications to run on most of the popular mobile platforms. As this layer is implemented based on Ionic framework, it also works on two kernel components, Apache Cordova and AngularJS. 1) Apache Cordova: Apache Cordova is an open-source mobile development framework. To avoid each mobile platforms’ native development language, it allows developer to use standard web technologies (such as, HTML5, CSS3, and JavaScript, etc.) for cross-platform development. Mobile applications are executed within wrappers on different platforms, and they rely on the standard APIs

to access the different device’s sensors, data, and network status. To maximize the native mobile devices hardware capabilities, the Apache Cordova framework, as a plugin, is integrated into the new application architecture. 2) AngularJS: AngularJS is a structural framework for dynamic web applications development. Developers can use HTML as the template language, and extend HTML’s syntax to express the application’s components clearly and succinctly. It also provides the data binding function and dependency injection function to eliminate the redundant code of the target mobile application. As more and more new APIs are constantly developed for the interactive features of mobile applications, the Ionic framework supports more and more mobile systems, e.g., Android, iOS, Windows Mobile, Blackberry, Amazon Fire OS, Firefox OS, Ubuntu Mobile OS, and Tizen. To enhance the mobile-optimized application performance, Crosswalk, as the rendering runtime engine, is integrated into the mobile-optimized application. Moreover, the Crosswalk also runs as a runtime engine in the different mobile systems to automatically update the rendering engine based on the different platforms. To develop an easy-touse application UI, the optimized application layer relies on standard APIs of WebView for the application presentation layer. As the WebView can effectively improve the performance and user experience, it works as a middleware between the Web technology (such as, AngularJS, HTML5, Apache Cordava) and native mobile systems. Depending on the different framework (such as, AngularJS, Apache Cordava), many different type WebViews can be used for mobile application implementation. The detailed modules of the optimized application layer are listed as follows. 1) Apache Cordova module. 2) Web application support module (implemented with AngularJS, Ionic framework, and common codebase). 3) WebView (Crosswalk Rendering runtime engine). 4) Mobile systems native APIs module (Android, iOS, Windows Mobile, etc.). 2) Unified Framework Layer: In order to integrate the remote laboratory technology into the new mobile-optimized application architecture, the unified framework is merged into


this new mobile application framework. The unified framework is based on the combination solution of both Apache web engine and Node.js web engine, and it uses a Node-HTTP-based technology to implement the real-time communication between experiment equipment with end users. The subsequent iteration of the design resolved challenges of developing cross-browser and cross-device web UI as an improvement to the unified framework. To implement the real-time communication between experiment equipment with mobile application without plug-ins, a real-time communication module based on LtoN protocol and an experiment control program are developed using LabVIEW. National Instrument’s LabVIEW software is one of the popularly deployed technologies for remote panel over Internet [54]. With the development of computer technology, LabVIEW also integrated a new feature to interact with the experiment virtual instruments by using RESTful web services technology. REST (representational state transfer) provides a lightweight protocol accessible to a wide variety of clients. The architecture does not require complex message passing and provides a simple interface for user to begin using Web services in LabVIEW. However, it requires the client interface to be developed using different technologies, and LabVIEW plug-ins must be installed in the web browsers [55]. To resolve this plug-in issue and support the new mobile-optimized remote laboratory application architecture, the new application communication module was implemented based on the LtoN protocol. This module includes three parts, a client part that runs in mobile-optimized application, a server part runs in the Node.js web server and a control module runs in LabVIEW experiment equipment control program. Client part and server part were developed with JavaScript language and the control module was developed in LabVIEW. With this new application communication module, the mobileoptimized application can real-time communicate with experiment equipment without LabVIEW plug-ins in client side. The detailed modules of unified framework layer are listed as follows. 1) Real-time video transmission module (implemented based on HLS protocol). 2) Real-time experiment data transmission module (implemented based on LtoN protocol based on web socket). 3) Combinational server engine (Node.js web server and Apache Web server) 4) Common codebase (HTML, JavaScript, CSS). 5) Data management system (MySQL). 6) Experiment equipment control module (implemented by LabVIEW VIs). B. Characteristics of the New Mobile-Optimized Application Architecture Table IV depicts the characteristics of the new mobileoptimized application architecture. As the Apache Cordova framework supports cross-platform application development, the mobile-optimized architecture can support most of the popular mobile systems. In addition, since the AngularJS framework can support dynamic web applications development, lots of native features of different mobile systems can be realized based

2387

TABLE IV CHARACTERISTIC OF THE REMOTE EXPERIMENT OPTIMIZED APPLICATIONS Characteristic 1. Cross-platform 2. Native-Like features 3. Code reusable 4. Fragmented OS compatible 5. High performance 6. User Friendly 7. Real-time experiment 8. Global firewall transmission

Fig. 4.

Technology Apache Cordova Apache Cordova/AngularJS HTML/JavaScipt/CSS/WV (WebView) Crosswalk Runtime Engine Ionic Framework/AngularJS Ionic Framework The Unified Framework The Unified Framework

Equipment of the PID motor control experiment.

on the new mobile-optimized application architecture. WebView is also the key technology in this new architecture to support reusing the common codebase. Therefore, developers can deploy the mobile-optimized applications without spending much time on rewriting the code for different mobile platforms based on the new application architecture. IV. IMPLEMENTATION OF THE NEW APPLICATION An ME experiment, PID motor speed control experiment, has been incorporated as a part of the remote laboratory series used in the Mechanics, Controls and Vibrations Laboratory (MCVL) course at the University of Houston. In the MCVL course, the assignment for the remote PID control of a DC motor includes two labs. In the first lab, the students are asked to design three sets of experiments to clearly show the characteristics of the P controller, I controller, and D controller based on the knowledge they have learned from the other control theory courses. In the second lab, the students are required to estimate the values of the damping ratio, gain, and natural frequency of a second-order system model of the open-loop DC motor system. Based on the values estimated, the model of the DC motor is built with MATLAB/Simulink and verified with the experiment data. For the remote experiment integration, the detail process is given below. A. Experiment Hardware Setup As shown in Figs. 4 and 5, a remote PID motor controller experiment is built to help students to understand the characteristics of proportional, proportional-integral proportionalderivative, and PID controllers and visualize the process of remote tuning. The experiment shows student the principle of dynamic systems analysis as well as how to achieve the

2388

Fig. 5.


Close-up of dc motor.

system optimal behaviors depending on different applications. The controller is designed to control the angular velocity of a DC motor to follow various input signals (i.e., sinusoidal, triangular, or square waves) at frequencies of 0.1 to 1 Hz. Both the input and output angular velocity can be adjusted and displayed in real time, while the rotation of motor will be able to observe through a webcam in the remote laboratory. Saved data include input angular velocity, measured output angular velocity, and time in seconds. The remote PID motor control hardware, as shown in Fig. 4, contains: 1) DC motor with tachometer; 2) power amplifier; 3) power supply for the amplifier; 4) NI USB-6361 X Series DAQ; 4) PC workstation. The NI USB-6361 X Series DAQ is utilized to both measure the voltage signal from tachometer as the rotation speed of DC motor and generate analog output voltage as control signal. The power amplifier and power supply provide the electronic power to drive the DC motor rotating. In this experimental setup, the LabVIEW VIs, which execute on the PC workstation, are used to control the motor. During the experiment, LabVIEW generates the reference signal, such as square wave or triangle wave at designated amplitude and frequency. Based on the reference signal and the feedback DC motor speed signal, the PID VIs compute the control signal applied on the DC motor. To optimize the performance of the system, all the P, I, and D parameters can be tuned both locally and remotely. B. Experiment Software Integration For the new mobile-optimized remote PID motor speed control experiment software integration, three tasks (i.e., the optimized application layer implementation, the unified framework layer integration, and integrate the LtoN module to the new application architecture) have to be done. 1) Mobile-Optimized Remote Experiment Application Implementation: To implement the mobile-optimized remote experiment application, three key technologies, the Ionic framework, the Crosswalk runtime engine, and the Apache Cordova framework, have to be used. With Web 2.0 technology, Ionic framework uses the common codebase, which includes HTML, CSS, and AngularJS/JavaScript. In the application

Fig. 6. Block diagram of the new mobile-optimized remote PID motor speed control experiment user interface.

implementation process, the common codebase is wrapped into the Cordova framework and is rendered in the UI layer with the Crosswalk runtime engine. Normally, the common codebase is reusable to deploy the mobile-optimized application to the different mobile platforms. The Crosswalk works as a middleware to connect the Ionic framework with the unified framework. With these three key technologies, all of the source code and software plugins are packaged with a new mobile-optimized remote experiment application. To package the new mobileoptimized application, the Eclipse, a popular integrated development environment is used. The mobile-optimized remote PID motor speed control experiment application UI on three popular mobile operating systems is shown in Fig. 6. The UI includes three parts: 1) experiment real-time video; 2) real-time experiment data display; 3) experiment control components. With the server-based Mashup technology, the data are analyzed and reformatted on the server side, and then the data are transmitted to the users’ mobile Crosswalk rendering runtime engine. The architecture of the new client side rendering scheme is divided into three parts. 1) Presentation/Interaction: The new mobile-optimized application architecture uses the Crosswalk, as a rendering runtime engine, to render the UI WebView. 2) Web Services: The system functionality can be accessed using the API services. Three protocols, JSON-RPC, REST, and SOAP are used. 3) Real-time data transmission: The data are handled in three ways, sending, storing, and receiving. JSON and Socket.IO are used for data transmission. 2) Unified Framework Layer Integration: The unified framework layer is directly built on the top of an assembled server engine scheme. It includes two server engines working together, Apache HTTP server engine and Node.js server engine. With the server-based Mashup technology, the Apache HTTP server engine is used to combine the UI widgets and web content (such as, the real-time experiment data, the real-time experiment video) together. Meanwhile, the Node.js web engine handles the real-time experiment data transmission. The experiment scheduler system and user management system are also integrated into the unified framework. For the data management, the MySQL 5.5 database management system is used.


3) Integrate the LtoN Module to the New MobileOptimized Application Architecture: A real-time experiment data transmission protocol, named LtoN, is designed and developed based on Socket.IO protocol. With the new real-time transmission protocol, students can conduct the experiment, and save the experiment data. To integrate the LtoN protocol to the new mobile-optimized application architecture, some improvements need to be implemented. More details of the improved LtoN protocol are illustrated in the followings. 1) The new application communication protocol includes two parts, client part running in Crosswalk rendering runtime engine and server part running in web server. It is developed by AngularJS/JavaScript language and enhanced by the web socket protocol. 2) In this new application transmission protocol, we defined our own special communication instruction set to implement real-time experiment control commands and experimental data transmission. With the new communication instruction set, we can secure the data communication when user is conducting the remote experiment. To implement the improved LtoN protocol, only two JavaScript files need to be created. One for the client application is running in Crosswalk rendering runtime engine, and the other one for the server application is running in web server. Meanwhile, the new Node.js task needs to be created to run the protocol in server side, and this server-side application must hold running status forever. To ensure the normal real-time communication tasks, the server-side protocol application must held the active status. On the client side, a set file of communication instruction functions are used to support the normal operation of the remote experiment optimized application in Crosswalk rendering runtime engine. C. Performance of the Mobile-Optimized Remote PID Motor Speed Control Experiment Application To test the performance of the new mobile-optimized remote PID motor speed control experiment application, the Baidu mobile cloud testing center (http://mtc.baidu.com) is used. Baidu mobile cloud testing center provides free mobile application testing and porting service to any mobile application developers. After uploading the applications into the testing center, developers can test all functions of applications from startup to shut down, and once testing is complete the test center will provide a test report. The mobile-optimized application has been tested via automated testing. It has over 10 000 mobile devices which cover 1500 different mobile platforms. Automated testing uses scripting environment, which we can invoke the application functionality via scripts (for example, Python) to customize some tests, such as phone calls, send text messages, browse the web. The testing center can expand mobile application API, and calls the API in Python scripts, enabling extensive testing. To compare the performance with native remote experiment, we found an open source native remote experiment application namely RoboLiterate. RoboLiterate is the all-in-one Bluetooth remote control experiment to control Lego Mindstorms NXT

2389

TABLE V PERFORMANCE TEST RESULT Test items

CPU occupancy rate Memory usage Package size Starting duration Content loading time Charge depleting Frame rate test

Optimized remote control experiment application (PID motor speed control)

Native remote control experiment application (RoboLiterate)

2.66667% 53322 KB 1459 KB 321 ms 1010 ms 167.234 mA 3.28 fps

5.68% 18847.55 KB 4.5 MB 506.21 ms 1010 ms 0.15 mA 2.68 fps

robot. Table V depicts the test result. In this test, we compared the mobile-optimized remote PID experiment application with a native remote control experiment application, RoboLiterate. Because these two applications have a big difference in package size, 1459 KB versus 4.5 MB, the mobile device CPU occupancy rate, memory usage, starting duration are also different. As we know, the key evaluation metric for mobile application is the content loading time. From the result, these two applications’ content loading time is the same. Consequently, the mobileoptimized remote PID experiment application has native-like performance as it blends some advantages of native functions of mobile devices and some mobile web browser capabilities for M-Learning. V. FUTURE WORKS Although the new mobile-optimized remote laboratory application architecture delivers a new development tool to support student-centered M-Learning, there are still further development required to improve the new architecture stability and usability. More specifically, issues that need improvement are as follows. 1) Integrating the new mobile-optimized remote laboratory application architecture into the learning management systems (LMS). Currently, we developed our own remote laboratory management platform, and it includes a scheduler, a user management module, and a learning materials management module. In future, we plan to integrate our mobile-optimized remote laboratory application architecture into an open source LMS (e.g., Moodle) to avoid double registration of students. 2) Cloud computing technology will be used to more new mobile-optimized remote experiment applications deployment in future. The new mobile-optimized remote laboratory application architecture is still version 1.0. So far, only a few sample pilot tests were conducted. Some bugs in the software package have been fixed through students’ feedback. More comprehensive testing and user survey need to be done. VI. CONCLUSION In this paper, a new mobile-optimized application architecture was designed and implemented successfully to provide a

2390


new tool for M-Learning application and industrial electronics application development. The new mobile-optimized application architecture integrated the advantages of both native mobile applications and web applications. It improved the running performance and solved hardware accessibility issue of web applications. It also solved the cross-platform running issue of native application. Moreover, it seamlessly combined the unified framework and Ionic framework together to deliver the excellent remote laboratory services to students. As a pilot M-Learning application based on the new architecture, a new mobile remote PID experiment application was implemented successfully. It can connect the students to the real PID motor speed control experiment through the M-Learning environment anywhere and anytime. The future work will be to further refine and improve this mobile-optimized application architecture for M-Learning. ACKNOWLEDGMENT The authors would like to thank the anonymous referees for their helpful comments and suggestions. REFERENCES [1] G. Papagiannakis, G. Singh, and N. Magnenat-Thalmann, “A survey of mobile and wireless technologies for augmented reality systems,” Comput. Animation Virtual Worlds, vol. 19, no. 1, pp. 3–22, Mar. 2008. [2] M. Saylor, “Active learning and virtual worlds,” in The Mobile Wave: How Mobile Intelligence Will Change Everything. New York, NY, USA: Vanguard, Jan. 2013, pp. 176–180. [3] G. I. Analysts, “ELearning (MCP-4107): A global strategic business report,” Global Ind. Analysts, Inc., San Jose, CA, USA, Mar. 2014, pp. 9–11. [4] A. Kukulska-Hulme, “Smart devices or people? A mobile learning quandary,” Int. J. Learn. Media, vol. 4, no. 3/4, pp. 73–77, Aug. 2012. [5] M. Ally and J. Prieto-Blázquez, “What is the future of mobile learning in education?,” Revista Univ. Soc. Conocimiento, vol. 11, no. 1, pp. 142–151, Jan. 2014. [6] C. Wen and J. Zhang, “Design of a microlecture mobile learning system based on smartphone and web platforms,” IEEE Trans. Educ., vol. 58, no. 3, pp. 203–207, Aug. 2015. [7] J. S. Sung, “Design of smart learning in mobile environment,” Int. J. Softw. Eng. Appl., vol. 9, no. 12, pp. 373–380, Dec. 2015. [8] W. H. Wu, Y. C. J. Wu, C. Y. Chen, H. Y. Kao, C. H. Lin, and S. H. Huang, “Review of trends from mobile learning studies: A meta-analysis,” Comput. Educ., vol. 59, no. 2, pp. 817–827, Sep. 2012. [9] L. Gomes and S. Bogosyan, “Current trends in remote laboratories,” IEEE Trans. Ind. Electron., vol. 56, no. 12, pp. 4744–4756, Dec. 2009. [10] V. J. Harward et al., “The iLab shared architecture: A web services infrastructure to build communities of internet accessible laboratories,” Proc. IEEE, vol. 96, no. 6, pp. 931–950, Jun. 2008. [11] P. Orduna, J. Irurzun, L. Rodriguez-Gil, J. Garcia-Zubia, F. Gazzola, and D. Lopez-de-Ipina, “Adding new features to new and existing remote experiments through their integration in WebLab-Deusto,” Int. J Online Eng., vol. 7, no. S2, pp. 33–39, Jun. 2011. [12] W. S. Hu, G. P. Liu, and H. Zhou, “NCSLab: A web-based global-scale control laboratory with rich interactive features,” IEEE Trans. Ind. Electron., vol. 57, no. 10, pp. 3253–3265. Oct. 2010 [13] Y. L. Qiao, G. P. Liu, G. Zheng, and W. S. Hu, “ Web-based 3-D control laboratory for remote real-time experimentation,” IEEE Trans. Ind. Electron., vol. 60, no. 10, pp. 4673–4682, Oct. 2013. [14] A. Melkonyan, A. Gampe, M. Pontual, and G. Huang, “Facilitating remote laboratory deployments using a relay gateway server architecture,” IEEE Trans. Ind. Electron., vol. 61, no. 1, pp. 477–485, Jan. 2014. [15] A. A. Kist, P. Gibbings, A. D. Maxwell, and H. Jolly, “Supporting remote laboratory activities at an institutional level,” Int. J. Online Eng., vol. 9, no. S5, pp. 38–47, Oct. 2013. [16] D. May, C. Terkowsky, T. Haertel, and C. Pleul, “Bringing remote labs and mobile learning together,” Int. J. Interactive Mobile Technol., vol. 7, no. 3, pp. 54–62, Sep. 2013.

[17] J. B. Silva, W. Rochadel, J. P. Simão, R. Marcelino, and V. Gruber, “Using mobile remote experimentation to teach physics in public school,” in Proc. Int. Conf. Interactive Comput. Aided Blended Learn., 2013, pp. 46–51. [18] S. A. Saleh and B. S. Ahmad, “Mobile learning: A systematic review,” Int. J. Comput. Appl., vol. 114, no. 11, pp. 1–5, Nov. 2015. [19] H. Crompton, D. Burke, K. H. Gregory, and C. Gräbe, “The use of mobile learning in science: A systematic review,” J. Sci. Educ. Technol., vol. 25, no. 2, pp. 149–160, Apr. 2016. [20] D. Ivanc, R. Vasiu, and M. Onita, “Usability evaluation of a LMS mobile web interface,” Information and Software Technologies, Berlin, Germany: Springer, Oct. 2013, pp. 348–361. [21] S. Teri, A. Acai, D. Griffith, Q. Mahmoud, D. W. Ma, and G. Newton, “Student use and pedagogical impact of a mobile learning application,” Biochem. Mol. Biol. Educ., vol. 42, no. 2, pp. 121–135, Apr. 2014. [22] A. Juntunen, E. Jalonen, and S. Luukkainen, “Html 5 in mobile devices— Drivers and restraints,” in Proc. 46th IEEE Hawaii Int. Conf. Syst. Sci., 2013, pp. 1053–1062. [23] A. Charland and B. Leroux, “Mobile application development: Web vs native,” Commun. ACM, vol. 54, no. 5, pp. 49–53, May 2011. [24] B. Deaky, D. G. Zutin, and P. H. Bailey, “The first android client application for the iLab shared architecture,” Int. J. Online Eng., vol. 8, no. 1, pp. 4–7, Mar. 2012. [25] A. Hossain, J. Canning, S. Ast, P. J. Rutledge, T. L. Yen, and A. Jamalipour, “Lab-in-a-phone: Smartphone-based portable fluorometer for pH measurements of environmental water,” IEEE Sensors J., vol. 15, no. 9, pp. 5095–5102, Sep. 2015. [26] A. Alkouz, A. Y. Al-Zoubi, and O. Mohammed, “J2ME-based mobile virtual laboratory for engineering education,” Int. J. Interactive Mobile Technol., vol. 2, no. 2, pp. 5–10, Jun. 2008. [27] D. Chaos, J. Chacón, J. A. Lopez-Orozco, and S. Dormido, “Virtual and remote robotic laboratory using EJS, MATLAB and LabVIEW,” Sensors, vol. 13, no. 2, pp. 2595–2612, Feb. 2013. [28] C. A. Garc and P. Merino, “Remote control and instrumentation of Android devices,” in Proc. IEEE Int. Conf. Remote Eng. Virtual Instrum. 2016, pp. 190–195. [29] C. Onime and O. Abiona, “3D mobile augmented reality interface for laboratory experiments,” Int. J. Commun. Netw. Syst. Sci., vol. 9, no. 4, pp. 67–76, Apr. 2016. [30] J. Garcia-Zubia, D. López-de-Ipiña, and P. Orduña, “Mobile devices and remote labs in engineering education,” in Proc. 8th IEEE Int. Conf. Adv. Learn. Technol., Jul. 2008, pp. 620–622. [31] I. Gustavsson et al., “On objectives of instructional laboratories, individual assessment, and use of collaborative remote laboratories,” IEEE Trans. Learn. Technol., vol. 2, no. 4, pp. 263–274, Oct./Dec. 2009. [32] C. Terkowsky, C. Pleul, I. Jahnke, and A. E. Tekkaya, “Tele-operated laboratories for online production engineering education-platform for E-learning and telemetric experimentation (PeTEX),” Int. J. Online Eng., vol. 7, no. S1, pp. 37–43, Mar. 2011. [33] J. B. Silva, W. Rochadel, J. P. S. Simao, S. M. S. Bilessimo, and P. C. Nicolete, “Using mobile devices for conducting experimental practices in basic education,” Online Experimentation: Emerging Technologies and IoT, vol. 1, Castelldefels, Barcelona, Spain: IFSA, Dec. 2015, pp. 401– 418. [34] J. B. Ortega, E. B. Portas, J. A. L. Orozco, J. A. B. Seco, and J. M. Cruz, “Remote web-based control laboratory for mobile devices based on EJsS, raspberry Pi and node.js,” IFAC-PapersOnLine, vol. 48, no. 29, pp. 158–163, Nov. 2015. [35] F. Lustig, J. Dvorak, and P. Brom, “Simple modular system “iSES Remote Lab SDK" for creation of remote experiments accessible from PC, tablets and mobile phones: Workshop,” in Proc. IEEE Int. Conf. Remote Eng. Virtual Instrum., 2016, pp. 406–408. [36] N. Serrano, J. Hernantes, and G. Gallardo, “Mobile web apps,” IEEE Softw., vol. 30, no. 5, pp. 22–27, Sept./Oct. 2013. [37] N. Wang, X. Chen, G. Song, and H. Parsaei, “Remote experiment development using an improved unified framework,” in Proc. Int. Conf. E-Learn. World Conf. E-Learn. Corporate, Government, Healthcare, Higher Educ., 2014, pp. 2003–2010. [38] N. Wang, X. Chen, G. Song, and H. Parsaei, “Using node-HTTP-proxy for remote experiment data transmission traversing firewall,” Int. J. Online Eng., vol. 11, no. 2, pp. 60–67, Jun. 2015. [39] N. Wang, X. Chen, G. Song, and H. Parsaei, “A novel real-time video transmission approach for remote laboratory development,” Int. J. Online Eng., vol. 11, no. 1, pp. 1–4, Mar. 2015.


[40] J. Garc´ıa-Zubia, P. Orduˇna, D. López-de-Ipiˇna, and G.R. Alves, “Addressing software impact in the design of remote laboratories,” IEEE Trans. Ind. Electron., vol. 56, no. 12, pp. 4757–4767, Dec. 2009. [41] M. Tawfik, D. Lowe, C. Salzmann, D. Gillet, E. Sancristobal, and M. Castro, “Defining the critical factors in the architectural design of remote laboratories,” IEEE Rev. Iberoamer. Tecnol. Aprendizaje, vol. 10, no. 4, pp. 269–279, Nov. 2015. [42] R. T. Fielding and G. Kaiser, “The apache HTTP server project,” IEEE Internet Comput., vol. 1, no. 4, pp. 88–90. Jul./Aug. 1997. [43] D. Herron, “The capabilities of node,” Node Web Development, 2nd ed., Birmingham, U.K.: Packt, Jul. 2013, pp. 8–9. [44] R. Rai, “The Socket.IO protocol,” Socket.IO Real-Time Web Application Development. Sebastropol, CA, USA: O’Reilly Media, Feb. 2013, pp. 87–91. [45] X. Chen, D. Osakue, N. Wang, H. Parsaei, and G. Song, “Development of a remote experiment under a unified remote laboratory framework,” in Proc. Int. Conf. World Congr. Eng. Educ., 2013. [46] I. Dalmasso, S. K. Datta, C. Bonnet, and N. Nikaein, “Survey, comparison and evaluation of cross platform mobile application development tools,” in Proc. 9th. Int. Wireless Commun. Mobile Comput. Conf., 2013, pp. 323–328. [47] S. Ottka, “Comparison of mobile application development tools for multiplatform industrial applications,” Master’s thesis, Degree Progr. Comput. Sci. Eng., Aalto Univ., Espoo, Finland, 2015. [48] R. M. Babu, M. B. Kumar, R. Manoharan, M. Somasundaram, and S. P. Karthikeyan, “Portability of mobile applications using phonegap: A case study,” in Proc. Int. Conf. Softw. Eng. Mobile Appl. Modelling Develop., 2012, pp. 1–6. [49] H. Heitkötter, T. A. Majchrzak, B. Ruland, and T. Weber, “Evaluating frameworks for creating mobile web apps,” in Proc. 9th Int. Conf. Web Inf. Syst. Technol.,2013, pp. 209–221. [50] M. Ciman, O. Gaggi, and N. Gonzo, “Cross-platform mobile development: A study on apps with animations,” in Proc. 29th Annu. ACM Symp. Appl. Comput. Conf., 2014, pp. 757–759. [51] Research2guidance, “Cross platform tool benchmarking 2013 hidden champions of the app economy,” Tech. Rep., [Online]. Available: http://www.research2guidance.com/r2g/Cross-Platform-ToolBenchmarking-2013.pdf, Oct. 2013. [52] A. Biharisingh,“Build your first mobile app with the ionic framework— part,” [Online]. Available: http://gonehybrid.com/build-your-first-mobileapp-with-the-ionic-framework-part-1, Jan. 2015 [53] R. Khanna and M. Harlington, “Anatomy of a hybrid mobile app,” Getting Started With Ionic, Birmingham, U.K.: Packt, Jan. 2016, pp. 4–5. [54] N. Duro et al., “An integrated virtual and remote control lab: The threetank system as a case study,” Comput . Sci . Eng., vol. 10, no. 4, p. 50–59, Jul./Aug. 2008. [55] P. Orduña, J. Garc´ıa-Zubia, L. Rodriguez-Gil, J., Irurzun, D. López-deIpiña, and F. Gazzola, “Using LabVIEW remote panel in remote laboratories: Advantages and disadvantages,” in Proc. IEEE Global Eng. Educ. Conf., 2012, pp. 1–7.

Ning Wang (SM’14) was born in Gansu, China. He received the B.S. degree in information management systems from the Department of Information Management Systems, China Agriculture University, Beijing, China, in 2002, the M.S. degree in software management science from Hong Kong Polytechnic University, Hong Kong, in 2008, and the M.S. degree in computer science from Texas Southern University, Houston, TX, USA, in 2014. He is currently working toward the Ph.D. degree in electrical engineering in the Department of Electrical and Computer Engineering, University of Houston, Houston, TX, USA. He has published more than 6 peer-reviewed journal papers and more than 13 conference papers. His research interests include remote laboratories, remote control, network technology, electrical control, and embedded system software design. Mr. Wang received the COSET Distinguished Graduate Student Award from Texas Southern University in 2014.

2391

Xuemin Chen (M’99–SM’08) was born in Jiangsu, China. He received the B.S., M.S., and Ph.D. degrees in electrical engineering from the Nanjing University of Science and Technology (NJUST), Nanjing, China, in 1985, 1988, and 1991, respectively. In 2006, he joined Texas Southern University, Houston, TX, USA, where he is currently an Associate Professor in the Engineering Department. From 1998 to 2006, he was a Postdoctoral Fellow and then a Research Assistant Professor in the Electrical and Computer Engineering Department, University of Houston. From 1991 to 1998, he was a faculty member in the Department of Automation at NJUST. He initiated the Virtual and Remote Laboratory (VR-Lab) in 2008 and served as the Founding Director of the VR-Lab at TSU since then. Gangbing Song (S’93–M’96) was born in Wuhan, China. He received the B.S. degree in energy engineering from Zhejiang University, Hangzhou, China, and the M.S. and Ph.D. degrees in mechanical engineering from the Department of Mechanical Engineering, Columbia University, New York, NY, USA, in 1989, 1991, and 1995, respectively. He is the Founding Director of the Smart Materials and Structures Laboratory and a Professor of mechanical, civil, and environmental engineering and electrical and computer engineering at the University of Houston, Houston, TX, USA, where he holds the John and Rebecca Moores Professorship. He has published more than 400 papers, including 200 peer-reviewed journal articles. He is also an inventor or coinventor of 11 U.S. patents and 11 pending patents. Dr. Song is a member of the American Society of Civil Engineers and American Society of Mechanical Engineers. Qianlong Lan was born in Guangxi, China, in 1991. He received the B.S. degree in network engineering from the Shanghai Second Polytechnic University, Shanghai, China, in 2013. He has been working toward the M.S. degree in computer science at Texas Southern University, Houston, TX, USA, since 2014. In 2014, he joined the Department of Computer Science, Texas Southern University as a Research Assistant. He has published five conference papers in his two years of study at Texas Southern University. Mr. Lan received the COSET Research Enrichment Scholarship from Texas Southern University for his contributions to the field of research in 2016. Hamid R. Parsaei was born in Tehran, Iran. He received the B.S. degree in economics from the National University of Iran, Tehran, Iran, and the M.S. and Ph.D. degrees in industrial engineering from Western Michigan University, Kalamazoo, MI, USA, in 1980, and The University of Texas at Arlington, Arlington, TX, USA, in 1984. He is currently a Professor of mechanical engineering at Texas A&M University at Qatar, Doha, Qatar, and holds the title of Professor with tenure in the Department of Industrial and Systems Engineering at Texas A&M University, College Station, TX, USA. He was an Associate Dean for academic affairs at Texas A&M University at Qatar (2010–2014), a Professor and the Chair of the Department of Industrial Engineering, University of Houston (2001–2010), a Professor of Industrial Engineering at the University of Louisville (1986–2000), and the State University of New York in Utica (1984–1986). He has published more than 270 articles in peer-reviewed archival journals and conference proceedings. Prof. Parsaei is a Registered Professional Engineer in the State of Texas. He is a Fellow of the Institute of Industrial and Systems Engineers and the American Society for Engineering Education.