- Email: [email protected]
Development of an undergraduate course – Internet-based instrumentation and control
Computers & Education 49 (2007) 330–344 www.elsevier.com/locate/compedu
Development of an undergraduate course – Internet-based instrumentation and control q Hanqi Zhuang *, Salvatore D. Morgera Florida Atlantic University, Department of Electrical Engineering, 777 Glades Road, Boca Raton, FL 33496, USA Received 16 August 2005; accepted 17 August 2005
Abstract The objective, strategy, and implementation details of a new undergraduate course, Internet-based Instrumentation and Control, are presented. The course has a companion laboratory that is supported by the National Science Foundation and industry. The combination is offered to senior-level undergraduate engineering students interested in sensing, instrumentation, control, and web programming that want to learn more about the integration of these technologies for solving real-world engineering problems. The course will also be offered to gifted high school seniors with similar interests and can serve as a vehicle to attract them to engineering disciplines. Preliminary assessment of the first offering of the course is encouraging and has shown that the course has achieved success in helping students understand concepts and master basic technologies for developing Internet-based automatic systems. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Internet programming; Control over the internet; Networked control; Undergraduate education
1. Introduction The rapid growth of the Internet provides tremendous opportunities for Internet-based automation. For example, household electronic devices such as lights, appliances, climate-control systems, q *
This work is partially supported by a National Science Foundation Grant No. DUE-0127451. Corresponding author. Tel.: +1 561 852 0798; fax: +1 561 297 2336. E-mail address: [email protected] (H. Zhuang).
0360-1315/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.compedu.2005.08.001
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
331
Fig. 1. A typical web-based application.
and surveillance cameras can be linked to the Internet through wired or wireless networks (DuttaRoy, 1999). Another example is Internet-based tele-surgery. A spectacular demonstration in telesurgery was the recent transatlantic cholecystectomy performed between France and New York City. The paper by Butner and Ghodoussi (2003) described the system used and the steps leading up to the operation. It is predicted that Internet-connected home area networks will soon penetrate domestic life and Internet-based remote surgical procedures will link medical professionals in global practice. The Internet can also be used as the infrastructure for industrial applications. An Internet-based control, monitoring, and operation scheduling system for heating, ventilation, and air-conditioning (HVAC) systems was presented in Lin and Broberg (2002). A Web-based application (an example is illustrated in Fig. 1) has the following elements: Computers that serve as Web Servers, where HTML, scripting, and other programs reside. Computers that serve as Process Servers, where the programs that compute business logic reside. Process Servers connect to Web Servers on one side and process hardware on the other. Process hardware that gets the job done at the remote side. Databases that store information critical to the operation of the system. Client computers with which users access web sites. Client computers are connected to Web Servers through either the Internet (WAN/LAN) or corporate Intranets (LAN).1 In the configuration shown in Fig. 1, the Customer LAN/WAN is the backbone of the Internet and the Corporate LAN is the backbone of the Intranet. A client is connected to a web application, which resides in the Web Server, through the Internet. The web application then communicates with a control application, which resides in the process server, through the Intranet. The process server communicates with plants (phones in this case) through direct wiring. 1
LAN: local area network, WAN: wide area network.
332
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Due to the tremendous growth of the Internet over the past decade, professionals trained in aspects of Internet programming for Instrumentation and Control are in high demand. Materials that fit classroom teaching and laboratory experimentation for undergraduate students in this area are, however, very scarce. Researchers have been working in the area of remote control suitable for classroom teaching for a numbers of years. In 1997, Zimmer, Kadionik, and Danto (1997) proposed the concept of utilizing web-based instrumentation tools. The main advantages, according to the authors, are the facts that students can be exposed to new and advanced instruments that are not facilitated in pure, classical university courses and that laboratories that are normally closed during the night can offer ‘‘night services’’ for students who live across time zones (Zimmer et al., 1997). Tan, Lee, and Soh (2002) reported their experience in developing an Internet-based system to allow monitoring of important process variables from a distributed control system (DCS). The system is part of a desktop DCS, which enables users to access DCS over the Internet for the purpose of distance learning. The hardware and software design and implementation considerations for such a system are also documented. Recently, a special issue of the IEEE Controls Magazine dealt with the subject of Wireless Networks and Web-based Education (2003). Researchers reported their experiences in conducting control system courses in a networked environment. These range from hardware configurations to software architectures, including design and implementation details of control experiments that are suitable for Internet-based learning. From the above discussion and other references (Furuya, Kato, & Sekozawa, 2000; Guangcheng, Fei, Changhong, W, & Yufeng, 2003; Jeon, Kim, Kim, Cho, & Lee, 2001; Kato, Furuya, Tamano-Mori, Kaneko, & Nakano, 2001; Overstreet & Tzes, 1999; Park & Lee, 2001; Ramakrishnan et al., 2000; Yang, Tan, & Chen, 2002; Zhang, Chen, Ko, Chen, & Ge, 2001), we notice that researchers have focused on the development of remote control laboratories mainly for distance learning. In such a scenario, researchers and usually a small group of students under the supervision of a professor are engaged in developing locally, i.e., in a laboratory, an Internetbased control system. Such a system is then used in teaching a control course over the Internet. This approach allows students to operate on control systems in a remote setting; therefore, theoretically, students can access the laboratory 7 days a week and 24 hours a day. The emphasis of the present work is not distance learning. It is to design a technical elective course that trains undergraduate students to develop, rather than use, Internet-based automation systems. It is believed that this is the first time that such a course has been offered to undergraduate students. This course requires that the students master knowledge on sensing, control, and actuation, as well as Internet programming. This range of material is traditionally drawn from different engineering disciplines. The objectives pursued in this endeavor are to demonstrate that: (a) it is practical and feasible to offer engineering undergraduate students a course, Internet-based Instrumentation and Control, which provides a platform for students to design and implement their own Internetbased automation systems; (b) the proposed course can be effectively conducted with two integrated components: classroom lecturing and hands-on practice; and (c) the course is a valuable component for strengthening the learning outcome related to the ability of working in interdisciplinary teams, as required by the Accreditation Board for Engineering and Technology (2000).
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
333
2. Fitting to the curriculum The target group for this technical elective course is mainly students from the disciplines of Electrical and Computer Engineering. At Florida Atlantic University, Electrical Engineering and Computer Engineering are housed in two different departments. The approach presented in this paper is not, however, in any way limited by this particular arrangement and applies equally well to the situation in which there is a single Electrical and Computer Engineering Department. Naturally, students from the Department of Computer Engineering take more courses on the software side, in addition to basic Electrical Engineering courses (such as Electronic Circuits and Digital Circuits). On the other hand, students from the Department of Electrical Engineering have very little training in computer languages (at Florida Atlantic University, however, all students must take Introduction to C), although they have much more exposure to electronics, instrumentation, and control. A prerequisite for the course is a programming course, Introduction to C, which is usually taken by engineering students in their first or second year of study. After taking Introduction to C, students should possess basic skills in using computers and writing simple programs. Once students have these skills, learning Visual Basic is not difficult. For this purpose, a few lectures are provided on Visual Basic before teaching any web programming and instrumentation materials. Our experience indicates that students can pick up Visual Basic quickly by doing a number of selected computer examples. In this way, the students do not need to take a separate course on Visual Basic. Desired prerequisite courses also include Fundamentals of Engineering, which is a required course for all Engineering freshmen, and Electronics I with Laboratory, in which students learn basic circuits along with simple sensors and actuators. The vast majority of Electrical Engineering and Computer Engineering academic programs across the nation offer these prerequisite courses prior to a studentÕs senior year of study. After, or concurrently with, taking the Internet-based Instrumentation and Automation course, students usually complete a college wide capstone engineering design course sequence (Engineering Design I and II at Florida Atlantic University). The new course fits very well in the intersection of the two departmentsÕ curricula. It provides Computer Engineering students with much needed knowledge on sensors, actuators, and data acquisition tools. On the other hand, it greatly improves Electrical Engineering studentsÕ programming skills. The projects conducted in the class are integrated in nature; therefore, they serve as a rehearsal for more broad engineering design projects that students will complete in the Engineering Design I and II course sequence. When students graduate, they will go mostly to industries, and to a lesser extent, to various government agencies and graduate schools. The understanding of technologies offered in the presented course will be a valuable asset for these students in solving real-world problems, as the Internet has penetrated almost every aspect of our daily life. Students will be better prepared for undertaking design and implementation of projects in the networked world. The wide range of subjects covered in the course can result in potential problems. For instance, programming exercises assigned at the beginning of the class seem easy for students from Computer Engineering, while they may pose challenges for Electrical Engineering students. On the other hand, simple feedback control concepts can be confusing for students from Computer Engineering. Additional materials that provide background materials are essential for the smooth offering of the course. Balance between internet programming and instrumentation is a fine act
334
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
that, if carried out with thought and planning, will provide all students with valuable knowledge and skills.
3. Methods The curriculum development plan is presented in this section. The goal of the plan is to train engineering students in the design and implementation of Internet-based automation systems. 3.1. Development plan The Internet-based Instrumentation and Control course has a companion laboratory, which hosts 12 workstations that are inter-connected via the campus LAN. Each workstation consists of a PC, a data acquisition board, and other equipment. A Web Server and a Process Server run on every workstation. A couple of web cameras are installed in the laboratory. In terms of the lecture material, the course addresses the following topics:
Web Programming Environment Web Programming Tools Sensors and Actuators Data acquisition Monitoring Processes Over the Internet Control Processes Over the Internet Security and Fault Tolerance Case Studies
The course has three credits, with two for lecture and one for laboratory experimentation. The laboratory portion of the course consists of six experiments. Each experiment requires approximately two to three weeks of laboratory time. Students need to do some preparation before the laboratory and write a report upon completion of the laboratory. After the completion of the laboratory exercises, students have the option of either taking a final test or completing a project. The experiments and projects focus on essential aspects of Internet-based automation applications such as multi-tier architecture, object-oriented programming, client side and server side programming, database usage, and realization of control logic with proper hardware components. The following considerations are taken into account in the preparation of the course/laboratory development plan: The course/laboratory will be multidisciplinary in nature. In order to develop a web-based industrial application, students will need to master relevant knowledge from various engineering disciplines. More specifically, students will need to understand the interactions among client computers, Web Servers, Process Servers, and process plants. They will be required to implement software components that perform specific engineering tasks. Further, issues that are unique for web-based applications will be addressed in depth. These include unpredictable time delay, security, and fault tolerance required for a reliable system over the Web.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
335
Most of the process-related labs and projects will be open-ended, encouraging students to find more effective solutions. For instance, there are many important issues involved in applications. These issues cannot possibly be all solved in the course. Any attempt by students to address some of the issues is, however, encouraged in the class. The lectures and laboratory experiments will be highly correlated, to reinforce concepts students have learned in the classroom. This will be critical to student learning, as the laboratory is a companion part of the course. Experiments will be arranged in a sequential order as listed in Section 3.3. In this way, students will apply what they have learned to the later experiments. The laboratory will promote both individuality and team spirit. The experiments designed in the courses will enhance the studentÕs ability to solve problems independently, while the term projects will promote a spirit of teamwork. To achieve this, it is required that students perform laboratory experiments individually and term projects in groups. The laboratory will provide remote accessing capability. Students will be able to complete the programming assignments (mostly individual work) at home with a computer, while achieving the objective of testing their implementation using devices in the laboratory. This will be feasible because most of the web project development tools now allow developers to work on a project in either the master mode or the local mode. In the master mode, students must work on the Web Server computer. On the other hand, in the local mode, the students can work on any computer that has a network connection and a simplified development environment. A significant portion of the laboratories outlined later in this section can be conducted in the local mode. This will stretch the boundary of the laboratory beyond the university campus. The system that we have developed does not support distance collaborative work at this stage.
3.2. Laboratory equipment There are two basic setups for the laboratory experiments and projects conducted in the course. The first experimental setup is made up of a simple physical plant, a controller, and a computer; with reference to Fig. 1, these are in place of the phone extensions, process server, and web server. The physical plant consists of a Parallax Boe-Bot robot and a Parallax Basic Stamp II Board of Education (Fig. 2). The controller is implemented in the Basic Stamp II microcontroller mounted on the Board of Education. The exercises on this and the other setup are developed in the Visual Basic.NET environment, which is installed in the computer. This setup was chosen because Visual Studio is very powerful, user-friendly and free for universities, and Basic Stamp II microcontrollers are easy to learn and have considerable I/O functionality. One can certainly use other platforms to conduct similar experiments. For instance, one can use Java instead of Visual Basic for Internet programming, although the learning curve of Java is relatively steep. The second setup is mainly made up of an Educational Servo ES 151 (including a Servo Box and an Actuator Unit as seen in Fig. 3) manufactured by Feedback Co. and a data acquisition board from National Instruments (NI). Again, a number of manufacturers provide a spectrum of experimental plants for college students to conduct control experiments. The selection of an Educational Servo system was due to its ease of use and relatively low cost. The Actuator Unit is comprised of the following components: A pre-amplifier that accepts two voltage inputs at its sockets; a servo amplifier (not shown in the picture) that accepts the
336
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Fig. 2. Basic Stamp II and Boe-Bot robot.
pre-amplifierÕs output and uses this to control the power to the motor; a command input generator, that is the large calibrated disc to the left, and is attached to an internal potentiometer which allows a command voltage to be produced; a control circuit; a meter that displays the current being supplied to the motor; and a motor position disc that indicates the degree of agreement between the command and the motor position. The Educational Servo box contains two major components, namely a physical plant and sensor as shown in Fig. 3. The plant consists of a 24 V dc electric motor, a gearbox, a worm and wheel used to drive an output shaft, and a tachogenerator, which is attached to the motor and generates a voltage related to speed. The sensor is composed of an output disc and potentiometer. This disc is calibrated to indicate output shaft position (angular) and has stroboscope tracks that can be used to determine certain speeds. Behind the disc is a sensor that will give a voltage output that is related to the shaft position. The data acquisition board from NI provides a set of functions that control all of the National Instruments plug-in DAQ (data acquisition) devices for analog I/O, digital I/O, timing I/O, and others. The NI-DAQ software package has both high-level DAQ I/O functions for maximum ease of use, and low-level DAQ I/O functions for maximum flexibility and performance. Examples of
Fig. 3. Educational Servo ES 151.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
337
high-level functions are streaming data to disk or acquiring a certain number of data points. Examples of low-level functions are writing directly to the DAQ device registers or calibrating the analog inputs. 3.3. Laboratory activities Laboratory activities are divided into two types: laboratory exercises and projects. Suggested laboratory exercises are described below (http://webdev08.ee.fau.edu/feedbackcontrolweb/). Visual Basic – Console Programming. Visual Basic, a powerful, yet easy to learn programming language, is widely used by engineers and professionals. In this course, Visual Basic is mainly used to write software components for specific tasks. Since it will be used extensively in the entire course, in this and the next couple of experiments, students will study basics of Visual Basic programming. In this experiment, students will be familiar with the programming environment, which includes Toolbox, Resource, Solution Explorer, etc. They will be able to create a new project in Visual Basic.NET and write a simple Visual Basic program with correct data types, statements, and operations. In the later stage of this experiment, students are required to use all kinds of control statements to form complicated conditions and to write procedures for simple tasks. An example exercise is to write a program that computes total and monthly salary of a person. The program takes the hourly rate from a user, calculates the total and average monthly salary of the year, and then prints out the results. Students must write a function to calculate and return the total and average monthly salary after getting the hourly rate. Visual Basic – Windows Programming. In this experiment, students will first learn to create a Graphic User Interface (GUI). The Visual Basic.NET toolbox provides rich components. After working with the GUI, students will learn how to handle exceptions using a try-catch block. A homework assignment requires students to create two windows applications: a math tutorial for elementary school students and a tic-tac-toe game. Specific tasks requested by the user will be entered and execution results will be presented through a GUI. An individual procedure is used to handle each task. For instance, routines are written to fetch and save data from/to a file. Students are also required to write code to handle common exceptions. Later in another experiment, database techniques will be introduced to provide students with an alternate means for data management. Visual Basic – Web Programming. The objective of this experiment is to provide tools for students to implement a three-tier application. To this end, students will learn to create a web application using Visual Basic.NET. The web application will run under MicrosoftÕs Internet Information Server (IIS). Eventually, the web application will have a number of tiers. The front tier consists of a set of web pages, interacting with remote users through a web browser. The middle tier links the remote user to data sources and plants. The back tier, in our case, is a Process Server, which will be discussed in detail in a later experiment. The web pages created will be decorated with graphics/image files such as logos and photos of the laboratory. From the pages, remote users should be able to log in to the site, fetch documents, such as laboratory instructions, and navigate to other locations with ease. In the homework assignment, students will move the math tutorial and tic-tac-toe applications to the web. Technically, the students have to divide each task into two parts: the GUI part and the
338
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344 Actuator
Physical Plant
Sensor
(Servo Motor)
(Robot)
(Encoder)
Controller 2 (Basic Stamp 2)
Transceiver 2 Transceiver 1 Controller 1 (Basic Stamp 1)
GUI (Windows Application)
Fig. 4. A diagram for RF remote control.
algorithm part. The GUI part will be similar to that of a window application, except that it is implemented using web components and will be delivered to a remote user through a browser in this case. The algorithm part, also similar to the one in the windows application, will be coded in a program that runs at the server side. Basic Stamp II. In this experiment, students will learn to program Basic Stamp microcontrollers. Basic Stamp is a single-board microcontroller that runs the Parallax PBASIC language interpreter. The developer stores all the code in an EEPROM,2 which can also be used for data storage. The PBASIC language has easy-to-use commands for basic I/O, such as turning devices on or off, interfacing with sensors, etc. More advanced commands let the Stamp interface with other integrated circuits, communicate with each other, and operate in networks. In this experiment, students will learn how to program Basic Stamp II, build a simple plant (i.e., a robotic cart and some parallel LEDs) and write a PBASIC program to control the plant. Control the Robot with Wireless Links. Once students are familiar with Basic Stamp microcontrollers, they are then ready to implement a simple control system. First, they will design a GUI as a Windows application in Visual Basic. They then write code to communicate the control commands and feedback signals via two Basic Stamp microcontrollers, both of which are interfaced with a pair of transceivers, as shown in Fig. 4. Through the experiment, students will interpret the control loop as follows: A user initiates a command by clicking a button on the GUI. The Windows program behind the GUI sends the command to Basic Stamp 1 through a serial communications port using a communication protocol (e.g., RS 232). Basic Stamp 1 talks with Basic Stamp 2 through a RF (radio frequency) link (the distance is up to 200 feet). Basic Stamp 2 interprets the command and sends a control signal to the physical plant. The feedback signal from the plant can be sent back to the user through the path in reverse order. Robot control system via the internet. Once the robot is properly controlled locally, the entire routine developed for the system will be ready to be moved to the Web. In this experiment, 2 EEPROM: electronically erasable programmable read-only memory, which can be programmed repeatedly and can keep data even after the power is shut off.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
339
students will control the robot remotely using web applications with transceivers and Basic Stamps. They are required to interface the robot with the Process Server through the transceivers and perform communications between the Process Server and the Web Server. They will also program the web camera to send image sequences of its environment to the remote user, providing visual feedback to the user. A block diagram of the control strategy is given in Fig. 5. A large portion of the diagram is identical to that of Fig. 4, except the Internet part. A user enters a command through a GUI at the client side. A web browser will send the command to the Web Server that hosts the web application. The web application creates a TCP client socket that communicates with the TCP server socket residing at the Process Server. The Process Server does the same job as the Windows application shown in Fig. 4. The remaining control actions are identical to those shown in Fig. 4. As previously mentioned, students can either take a final test or develop a project after completing the experiments. Those who choose to do a term project can either create an Internet-based control system similar to the one shown in Fig. 6 or develop a control system of their choice with the instructorÕs approval. The GUI for the servo control in Fig. 6 includes PD parameter settings and other functionality that students might want to add on. Moreover, the control plant is monitored by a web camera that is installed in the laboratory. A remote user selects the control type and sets control parameters through the GUI. The control command is then sent to the plant via the control path similar to that given in Fig. 5, in which case a data acquisition board from National Instruments is used to replace the Basic Stamps and wireless link. The controller, implemented in the
Actuator
Physical Plant
Sensor
(Led, Servos)
(Led, Robot)
(Led)
Controller2 (Basic Stamp2)
Transceiver2
Transceiver1 Controller1 (Basic Stamp1)
Server Socket (Process Server)
Client Socket (Web Server)
GUI (Browser, Cli ent side)
Fig. 5. A diagram for Internet-based control.
340
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Fig. 6. A typical graphic user interface.
Process Server, generates a control signal and sends it to the servo motor, which in turn moves the disk to the desired location in a closed loop. 4. Assessment and discussion In the summer of 2003, this course was offered for the first time to 20 junior/senior engineering students after about a year of preparation. The course was conducted in a six-week period, including classroom instruction and laboratory exercises. Students completed the first three laboratory exercises individually. Due mainly to the limited number of Basic Stamp microcontrollers, transceivers and robots, they completed the remaining three exercises and the optional term projects in teams (each team had two students.) At the end of the term, a laboratory test or a term project was conducted in the class to judge the achievements of the students in the following topics: 1. 2. 3. 4.
Create windows and web applications. Program Basic Stamp II microcontrollers. Interface PC with microcontrollers. Create windows and web applications to control devices through microcontrollers.
About 60% of the students chose to conduct the term project, and the rest did the laboratory test. Among the students who did term projects, a couple of teams produced excellent results. One team constructed an Internet-based soil moisture control system and another, an Internet-based temperature control system. Both projects involved sensing, Internet programming, and control. As to the laboratory test, students were asked to: (a) create a web application with a simple GUI to allow users to enter commands that, say, turn on and off a LED which is interfaced with a Basic Stamp; (b) send the user commands to the Basic Stamp through a serial port of a personal computer; and (c) carry out the user command using the Basic Stamp. The test seems simple, but it contains most of the components of an Internet-based control system. About half of the students completed the test without much difficulty. The remaining students had problems during various steps of the test. They were given hints to overcome hurdles, with some penalties. At the end, all of the students completed the test.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
341
At the end of the class, a questionnaire was distributed to the 20 students. The answers to sample questions were compiled and are given below. The numbers in the brackets are the number of responses. Some students gave multiple answers. Question 1: What was your original intention of taking the class?
To control electrical devices via the Internet (5). To learn user-friendly computer languages such as Visual Basic (5). Need an elective (4). To learn how to program microcontrollers such as Basic Stamp (3). It sounds very interesting to topic (2). To learn computer interface (2). To have a class with practical application and hands-on experience (2).
Most students expressed a desire to explore and learn new things. A student stated that she would like to find out more about web-based control to see if she would be more interested in it for the field (work). There were a few students that just wanted to have an elective to fulfill graduation requirements. Question 2: Has the original objective been achieved after taking the class? All the students except one stated that their original objectives were met. Many indicated that this was an interesting class that opened up their minds. They also stated that the knowledge learned could be applied to other classes like Engineering Design I and II. A majority of the students pointed out, however, that six weeks were too short to cover the materials in depth and more time was needed. The student who answered no to the question also explained that he needed more time to digest the material. Question 3: What is your opinion on the method of alternating between lecturing and laboratory exercises? A vast majority of students favored the integration of lecturing with the laboratory exercises. They stated that the laboratory exercises provide much needed hands-on experience that reinforces the learning. Some students even suggested that the lectures should move to the laboratory. One student suggested that the material provided in the lecture should match more closely what was exercised in the laboratory. The only student who gave a negative reply to the question proposed to have the first half semester on the theory and the second half semester on the laboratory exercises. Question 4: What is your opinion on the laboratory equipment? Most of students thought the laboratory setup was good. Basic Stamp is a very user-friendly microcontroller, and Microsoft Visual Basic.NET is a very powerful, yet easy to learn, tool to create web applications. Some students complained that Basic Stamp educational boards and the robots used in the laboratory were not sturdy enough. Some stated that more computer stations were needed to allow everyone to conduct the exercises independently. Question 5: What is your opinion on the lecturing? Most of students gave positive responses to this question, though they have, in general, more favorable opinions on the laboratory exercises than the lectures. They like practical examples more than pure theoretical discussions. They prefer more simple demonstrations to show them step by step how things work. They want to also see more discussion on the equipment used in the laboratory.
342
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Question 6: What is your opinion on the laboratory exercises? Students had, in general, very favorable opinions on these laboratory exercises. They stated that the way to arrange the laboratory exercises from simple to more complex ones was thoughtful. The fact that the last laboratory was an integration of the previous laboratories was especially well received. Some students were very excited about the success of controlling a robot through a wireless link over the Internet. One student said that creating a website to control something was very educational. Question 7: Any other comments? Sample answers to this question are given below:
This should be one of the first courses students are given. Overall good educational class. Great tool for my senior design (Engineering Design I and II) project. Very good class. One of the few classes I have taken with a practical application. Time went by too fast. Good class. Nice to see some more current technologies in (other) classes. Best wishes next year.
Since the number of students that took the survey was relatively small, the results cannot be considered conclusive. We feel, however, that the survey provided us with useful information that is being used to improve the course for subsequent offerings. The performance of the students that had taken the Internet-based Instrumentation and Control (IIC) course was tracked to see how their performance fared in the Capstone Engineering Design I and II (ED I and ED II) courses. Table 1 lists the grades of these (20) students in comparison with the entire class (about 90 students). The grade scale is 0–4, with 4 being the highest. A simple hypothesis test is conducted with the data. The null assumption is ‘‘there is no performance difference between the two groups of students’’. Since the sample size is relatively small, a t-test is used in this study. The level of significance a is selected to be 0.05. The corresponding right-tail critical value Za is 1.645. The standard deviation s for both ED I and ED II for the first group is approximately 0.5. The t-statistic in this case is x l t¼ s ; pffiffi n where the sample mean x is given in the 2nd column of Table 1, the population mean l, is approximated by the class mean given in the 3rd column, the sample standard deviation s is 0.5, and n is 20. The computed t statistic for each case is given in the last column of Table 1. These are much smaller than the critical value 1.645; therefore the null hypothesis is rejected. Based on the sample, there is sufficient evidence at the 5% significance level to conclude that the students who had taken Table 1 Performance comparison of the controlled group with their peers ED I grade average ED II grade average
Statistics of students who had taken IIC
Statistics of the entire class
t Statistic
3.58 3.65
3.25 3.29
2.87 3.13
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
343
Fig. 7. Pablo Sanchez, a member of the FAU team, with its robot (Boca Raton News, FL, 4/15/2004).
the Internet-based Instrumentation and Control course achieved better results in the Engineering Design II course. In a robot competition sponsored by the American Nuclear Society in Spring 2004, Florida Atlantic University (FAU)Õs teams came out on top. FAU sent four teams of undergraduate engineering majors to the University of Florida campus for the competition. The task was to pick up three items of ‘‘nuclear waste’’ that had been dropped on the floor in a ‘‘hot room.’’ Each team had to use a robot to scoop up the waste items and contain them in a least amount of time. The FAU students did a superb job in the competition and took the top two spots. Four of these students had taken the Internet-based Instrumentation and Control course the previous semester. Their success was due to a number of factors, one of which we feel was the hands-on experience in microcontrollers, windows programming, and sensing and control that they gained in the Internet-based Instrumentation and Control course (Fig. 7). 5. Conclusion The primary objective of this paper is to describe the development of a course for undergraduate engineering students who are interested in learning and applying Internet programming technology to tasks in Internet-based control. Based on the positive response of the students who have taken the course, it is our belief that the course is a viable elective engineering course that fits Electrical and Computer Engineering curricula very well. Future work includes refining the course, writing a textbook for this unique course, disseminating the results to other engineering colleges, and offering a similar course to high school students who are interested in pursuing engineering degrees.
Acknowledgements Efforts made by our graduate students, Cristian Popescu, Qingmei Li, and Hesong (Harry) Huang in preparing this course are greatly appreciated. The support provided by the National Science Foundation for the laboratory development component of the course is gratefully acknowledged.
344
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
References ABET Engineering Criteria. (2000). Accreditation Board for Engineering and Technology (ABET). Butner, S. E., & Ghodoussi, M. (2003). Transforming a Surgical Robot for Human Telesurgery. IEEE Journal of Robotics and Automation, 19(5), 818–824. Dutta-Roy, A. (1999). Networks for Homes. IEEE Spectrum, 26–39. Furuya, M., Kato, H., & Sekozawa, T. (2000). Secure Web-based monitoring and control system. In Proceedings of the 26th annual conference on IEEE industrial electronics society, Vol. 4, pp. 2443–2448. Guangcheng, M., Fei, T., Changhong, W., & Yufeng, W. (2003). Web-based control service for self-organization of IDCS. In Proceedings of the IEEE conference on control applications, Vol. 1, pp. 1198–1202. Jeon, J. M., Kim, D. W., Kim, H. S., Cho, Y. J., & Lee, B. H. (2001). An analysis of network-based control system using CAN protocol. Proceedings of the IEEE international conference on robotics & automation, Vol. 4, pp. 3577– 3581. Kato, H., Furuya, M., Tamano-Mori, M., Kaneko, S., & Nakano, T. (2001). Risk analysis and secure protocol design for WWW-based remote control with operation – privilege management. In Proceedings of the IEEE international conference on systems, man & cybernetics, Vol. 2, pp. 1107–1112. Lin, A., & Broberg, H. L. (2002). Internet-based monitoring and controls for HVAC applications. IEEE Industry Applications Magazine, 8(1), 49–54. Overstreet, J. W., & Tzes, A. (1999). Internet-based client/server virtual instrument designs for real-time remote-access control engineering laboratory. In Proceedings of the American control conference, Vol. 2, pp. 1472–1476. Park, J. W., & Lee, J. M. (2001). Transmission modeling and simulation for internet-based control. In Proceedings of the 27th annual conference on IEEE industrial electronics society, Vol. 1, pp. 165–169. Ramakrishnan, V., Zhuang, Y., Hu, S. Y., Chen, J. P., Ko, C. C., Chen, B. M., et al. (2000). Development of a webbased control experiment for a coupled tank apparatus. In Proceedings of the American control conference, Vol. 6, pp. 4409–4413. Tan, K. K., Lee, T. H., & Soh, C. Y. (2002). Internet-based Monitoring of Distributed Control Systems-An undergraduate Experiment. IEEE Transactions on Education, 45(2), 128–134. Wireless Networks and Web-based Education, a Special Issue. IEEE Control Systems Magazine 23(2), April 2003. Yang, S. H., Tan, L. S., & Chen, X. (2002). Requirements specification and architecture design for Internet-based control systems. In Proceedings of the computer software & applications conference, pp. 75–80. Zhang, J., Chen, J., Ko, C. C., Chen, B. M., & Ge, S. S. (2001). A web-based laboratory on control of a two degree of freedom helicopter. In Proceedings of the IEEE decision & control conference, Vol. 3, pp. 2821–2826. Zimmer, T., Kadionik, P., & Danto, Y. (1997). A World-Wide-Web based instrumentation pool real testing in a virtual world. In Proceedings of the 6th annual conference on IEEE industrial electronics society, Vol. 4, pp. 2443–2448.
Development of an undergraduate course – Internet-based instrumentation and control q Hanqi Zhuang *, Salvatore D. Morgera Florida Atlantic University, Department of Electrical Engineering, 777 Glades Road, Boca Raton, FL 33496, USA Received 16 August 2005; accepted 17 August 2005
Abstract The objective, strategy, and implementation details of a new undergraduate course, Internet-based Instrumentation and Control, are presented. The course has a companion laboratory that is supported by the National Science Foundation and industry. The combination is offered to senior-level undergraduate engineering students interested in sensing, instrumentation, control, and web programming that want to learn more about the integration of these technologies for solving real-world engineering problems. The course will also be offered to gifted high school seniors with similar interests and can serve as a vehicle to attract them to engineering disciplines. Preliminary assessment of the first offering of the course is encouraging and has shown that the course has achieved success in helping students understand concepts and master basic technologies for developing Internet-based automatic systems. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Internet programming; Control over the internet; Networked control; Undergraduate education
1. Introduction The rapid growth of the Internet provides tremendous opportunities for Internet-based automation. For example, household electronic devices such as lights, appliances, climate-control systems, q *
This work is partially supported by a National Science Foundation Grant No. DUE-0127451. Corresponding author. Tel.: +1 561 852 0798; fax: +1 561 297 2336. E-mail address: [email protected] (H. Zhuang).
0360-1315/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.compedu.2005.08.001
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
331
Fig. 1. A typical web-based application.
and surveillance cameras can be linked to the Internet through wired or wireless networks (DuttaRoy, 1999). Another example is Internet-based tele-surgery. A spectacular demonstration in telesurgery was the recent transatlantic cholecystectomy performed between France and New York City. The paper by Butner and Ghodoussi (2003) described the system used and the steps leading up to the operation. It is predicted that Internet-connected home area networks will soon penetrate domestic life and Internet-based remote surgical procedures will link medical professionals in global practice. The Internet can also be used as the infrastructure for industrial applications. An Internet-based control, monitoring, and operation scheduling system for heating, ventilation, and air-conditioning (HVAC) systems was presented in Lin and Broberg (2002). A Web-based application (an example is illustrated in Fig. 1) has the following elements: Computers that serve as Web Servers, where HTML, scripting, and other programs reside. Computers that serve as Process Servers, where the programs that compute business logic reside. Process Servers connect to Web Servers on one side and process hardware on the other. Process hardware that gets the job done at the remote side. Databases that store information critical to the operation of the system. Client computers with which users access web sites. Client computers are connected to Web Servers through either the Internet (WAN/LAN) or corporate Intranets (LAN).1 In the configuration shown in Fig. 1, the Customer LAN/WAN is the backbone of the Internet and the Corporate LAN is the backbone of the Intranet. A client is connected to a web application, which resides in the Web Server, through the Internet. The web application then communicates with a control application, which resides in the process server, through the Intranet. The process server communicates with plants (phones in this case) through direct wiring. 1
LAN: local area network, WAN: wide area network.
332
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Due to the tremendous growth of the Internet over the past decade, professionals trained in aspects of Internet programming for Instrumentation and Control are in high demand. Materials that fit classroom teaching and laboratory experimentation for undergraduate students in this area are, however, very scarce. Researchers have been working in the area of remote control suitable for classroom teaching for a numbers of years. In 1997, Zimmer, Kadionik, and Danto (1997) proposed the concept of utilizing web-based instrumentation tools. The main advantages, according to the authors, are the facts that students can be exposed to new and advanced instruments that are not facilitated in pure, classical university courses and that laboratories that are normally closed during the night can offer ‘‘night services’’ for students who live across time zones (Zimmer et al., 1997). Tan, Lee, and Soh (2002) reported their experience in developing an Internet-based system to allow monitoring of important process variables from a distributed control system (DCS). The system is part of a desktop DCS, which enables users to access DCS over the Internet for the purpose of distance learning. The hardware and software design and implementation considerations for such a system are also documented. Recently, a special issue of the IEEE Controls Magazine dealt with the subject of Wireless Networks and Web-based Education (2003). Researchers reported their experiences in conducting control system courses in a networked environment. These range from hardware configurations to software architectures, including design and implementation details of control experiments that are suitable for Internet-based learning. From the above discussion and other references (Furuya, Kato, & Sekozawa, 2000; Guangcheng, Fei, Changhong, W, & Yufeng, 2003; Jeon, Kim, Kim, Cho, & Lee, 2001; Kato, Furuya, Tamano-Mori, Kaneko, & Nakano, 2001; Overstreet & Tzes, 1999; Park & Lee, 2001; Ramakrishnan et al., 2000; Yang, Tan, & Chen, 2002; Zhang, Chen, Ko, Chen, & Ge, 2001), we notice that researchers have focused on the development of remote control laboratories mainly for distance learning. In such a scenario, researchers and usually a small group of students under the supervision of a professor are engaged in developing locally, i.e., in a laboratory, an Internetbased control system. Such a system is then used in teaching a control course over the Internet. This approach allows students to operate on control systems in a remote setting; therefore, theoretically, students can access the laboratory 7 days a week and 24 hours a day. The emphasis of the present work is not distance learning. It is to design a technical elective course that trains undergraduate students to develop, rather than use, Internet-based automation systems. It is believed that this is the first time that such a course has been offered to undergraduate students. This course requires that the students master knowledge on sensing, control, and actuation, as well as Internet programming. This range of material is traditionally drawn from different engineering disciplines. The objectives pursued in this endeavor are to demonstrate that: (a) it is practical and feasible to offer engineering undergraduate students a course, Internet-based Instrumentation and Control, which provides a platform for students to design and implement their own Internetbased automation systems; (b) the proposed course can be effectively conducted with two integrated components: classroom lecturing and hands-on practice; and (c) the course is a valuable component for strengthening the learning outcome related to the ability of working in interdisciplinary teams, as required by the Accreditation Board for Engineering and Technology (2000).
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
333
2. Fitting to the curriculum The target group for this technical elective course is mainly students from the disciplines of Electrical and Computer Engineering. At Florida Atlantic University, Electrical Engineering and Computer Engineering are housed in two different departments. The approach presented in this paper is not, however, in any way limited by this particular arrangement and applies equally well to the situation in which there is a single Electrical and Computer Engineering Department. Naturally, students from the Department of Computer Engineering take more courses on the software side, in addition to basic Electrical Engineering courses (such as Electronic Circuits and Digital Circuits). On the other hand, students from the Department of Electrical Engineering have very little training in computer languages (at Florida Atlantic University, however, all students must take Introduction to C), although they have much more exposure to electronics, instrumentation, and control. A prerequisite for the course is a programming course, Introduction to C, which is usually taken by engineering students in their first or second year of study. After taking Introduction to C, students should possess basic skills in using computers and writing simple programs. Once students have these skills, learning Visual Basic is not difficult. For this purpose, a few lectures are provided on Visual Basic before teaching any web programming and instrumentation materials. Our experience indicates that students can pick up Visual Basic quickly by doing a number of selected computer examples. In this way, the students do not need to take a separate course on Visual Basic. Desired prerequisite courses also include Fundamentals of Engineering, which is a required course for all Engineering freshmen, and Electronics I with Laboratory, in which students learn basic circuits along with simple sensors and actuators. The vast majority of Electrical Engineering and Computer Engineering academic programs across the nation offer these prerequisite courses prior to a studentÕs senior year of study. After, or concurrently with, taking the Internet-based Instrumentation and Automation course, students usually complete a college wide capstone engineering design course sequence (Engineering Design I and II at Florida Atlantic University). The new course fits very well in the intersection of the two departmentsÕ curricula. It provides Computer Engineering students with much needed knowledge on sensors, actuators, and data acquisition tools. On the other hand, it greatly improves Electrical Engineering studentsÕ programming skills. The projects conducted in the class are integrated in nature; therefore, they serve as a rehearsal for more broad engineering design projects that students will complete in the Engineering Design I and II course sequence. When students graduate, they will go mostly to industries, and to a lesser extent, to various government agencies and graduate schools. The understanding of technologies offered in the presented course will be a valuable asset for these students in solving real-world problems, as the Internet has penetrated almost every aspect of our daily life. Students will be better prepared for undertaking design and implementation of projects in the networked world. The wide range of subjects covered in the course can result in potential problems. For instance, programming exercises assigned at the beginning of the class seem easy for students from Computer Engineering, while they may pose challenges for Electrical Engineering students. On the other hand, simple feedback control concepts can be confusing for students from Computer Engineering. Additional materials that provide background materials are essential for the smooth offering of the course. Balance between internet programming and instrumentation is a fine act
334
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
that, if carried out with thought and planning, will provide all students with valuable knowledge and skills.
3. Methods The curriculum development plan is presented in this section. The goal of the plan is to train engineering students in the design and implementation of Internet-based automation systems. 3.1. Development plan The Internet-based Instrumentation and Control course has a companion laboratory, which hosts 12 workstations that are inter-connected via the campus LAN. Each workstation consists of a PC, a data acquisition board, and other equipment. A Web Server and a Process Server run on every workstation. A couple of web cameras are installed in the laboratory. In terms of the lecture material, the course addresses the following topics:
Web Programming Environment Web Programming Tools Sensors and Actuators Data acquisition Monitoring Processes Over the Internet Control Processes Over the Internet Security and Fault Tolerance Case Studies
The course has three credits, with two for lecture and one for laboratory experimentation. The laboratory portion of the course consists of six experiments. Each experiment requires approximately two to three weeks of laboratory time. Students need to do some preparation before the laboratory and write a report upon completion of the laboratory. After the completion of the laboratory exercises, students have the option of either taking a final test or completing a project. The experiments and projects focus on essential aspects of Internet-based automation applications such as multi-tier architecture, object-oriented programming, client side and server side programming, database usage, and realization of control logic with proper hardware components. The following considerations are taken into account in the preparation of the course/laboratory development plan: The course/laboratory will be multidisciplinary in nature. In order to develop a web-based industrial application, students will need to master relevant knowledge from various engineering disciplines. More specifically, students will need to understand the interactions among client computers, Web Servers, Process Servers, and process plants. They will be required to implement software components that perform specific engineering tasks. Further, issues that are unique for web-based applications will be addressed in depth. These include unpredictable time delay, security, and fault tolerance required for a reliable system over the Web.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
335
Most of the process-related labs and projects will be open-ended, encouraging students to find more effective solutions. For instance, there are many important issues involved in applications. These issues cannot possibly be all solved in the course. Any attempt by students to address some of the issues is, however, encouraged in the class. The lectures and laboratory experiments will be highly correlated, to reinforce concepts students have learned in the classroom. This will be critical to student learning, as the laboratory is a companion part of the course. Experiments will be arranged in a sequential order as listed in Section 3.3. In this way, students will apply what they have learned to the later experiments. The laboratory will promote both individuality and team spirit. The experiments designed in the courses will enhance the studentÕs ability to solve problems independently, while the term projects will promote a spirit of teamwork. To achieve this, it is required that students perform laboratory experiments individually and term projects in groups. The laboratory will provide remote accessing capability. Students will be able to complete the programming assignments (mostly individual work) at home with a computer, while achieving the objective of testing their implementation using devices in the laboratory. This will be feasible because most of the web project development tools now allow developers to work on a project in either the master mode or the local mode. In the master mode, students must work on the Web Server computer. On the other hand, in the local mode, the students can work on any computer that has a network connection and a simplified development environment. A significant portion of the laboratories outlined later in this section can be conducted in the local mode. This will stretch the boundary of the laboratory beyond the university campus. The system that we have developed does not support distance collaborative work at this stage.
3.2. Laboratory equipment There are two basic setups for the laboratory experiments and projects conducted in the course. The first experimental setup is made up of a simple physical plant, a controller, and a computer; with reference to Fig. 1, these are in place of the phone extensions, process server, and web server. The physical plant consists of a Parallax Boe-Bot robot and a Parallax Basic Stamp II Board of Education (Fig. 2). The controller is implemented in the Basic Stamp II microcontroller mounted on the Board of Education. The exercises on this and the other setup are developed in the Visual Basic.NET environment, which is installed in the computer. This setup was chosen because Visual Studio is very powerful, user-friendly and free for universities, and Basic Stamp II microcontrollers are easy to learn and have considerable I/O functionality. One can certainly use other platforms to conduct similar experiments. For instance, one can use Java instead of Visual Basic for Internet programming, although the learning curve of Java is relatively steep. The second setup is mainly made up of an Educational Servo ES 151 (including a Servo Box and an Actuator Unit as seen in Fig. 3) manufactured by Feedback Co. and a data acquisition board from National Instruments (NI). Again, a number of manufacturers provide a spectrum of experimental plants for college students to conduct control experiments. The selection of an Educational Servo system was due to its ease of use and relatively low cost. The Actuator Unit is comprised of the following components: A pre-amplifier that accepts two voltage inputs at its sockets; a servo amplifier (not shown in the picture) that accepts the
336
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Fig. 2. Basic Stamp II and Boe-Bot robot.
pre-amplifierÕs output and uses this to control the power to the motor; a command input generator, that is the large calibrated disc to the left, and is attached to an internal potentiometer which allows a command voltage to be produced; a control circuit; a meter that displays the current being supplied to the motor; and a motor position disc that indicates the degree of agreement between the command and the motor position. The Educational Servo box contains two major components, namely a physical plant and sensor as shown in Fig. 3. The plant consists of a 24 V dc electric motor, a gearbox, a worm and wheel used to drive an output shaft, and a tachogenerator, which is attached to the motor and generates a voltage related to speed. The sensor is composed of an output disc and potentiometer. This disc is calibrated to indicate output shaft position (angular) and has stroboscope tracks that can be used to determine certain speeds. Behind the disc is a sensor that will give a voltage output that is related to the shaft position. The data acquisition board from NI provides a set of functions that control all of the National Instruments plug-in DAQ (data acquisition) devices for analog I/O, digital I/O, timing I/O, and others. The NI-DAQ software package has both high-level DAQ I/O functions for maximum ease of use, and low-level DAQ I/O functions for maximum flexibility and performance. Examples of
Fig. 3. Educational Servo ES 151.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
337
high-level functions are streaming data to disk or acquiring a certain number of data points. Examples of low-level functions are writing directly to the DAQ device registers or calibrating the analog inputs. 3.3. Laboratory activities Laboratory activities are divided into two types: laboratory exercises and projects. Suggested laboratory exercises are described below (http://webdev08.ee.fau.edu/feedbackcontrolweb/). Visual Basic – Console Programming. Visual Basic, a powerful, yet easy to learn programming language, is widely used by engineers and professionals. In this course, Visual Basic is mainly used to write software components for specific tasks. Since it will be used extensively in the entire course, in this and the next couple of experiments, students will study basics of Visual Basic programming. In this experiment, students will be familiar with the programming environment, which includes Toolbox, Resource, Solution Explorer, etc. They will be able to create a new project in Visual Basic.NET and write a simple Visual Basic program with correct data types, statements, and operations. In the later stage of this experiment, students are required to use all kinds of control statements to form complicated conditions and to write procedures for simple tasks. An example exercise is to write a program that computes total and monthly salary of a person. The program takes the hourly rate from a user, calculates the total and average monthly salary of the year, and then prints out the results. Students must write a function to calculate and return the total and average monthly salary after getting the hourly rate. Visual Basic – Windows Programming. In this experiment, students will first learn to create a Graphic User Interface (GUI). The Visual Basic.NET toolbox provides rich components. After working with the GUI, students will learn how to handle exceptions using a try-catch block. A homework assignment requires students to create two windows applications: a math tutorial for elementary school students and a tic-tac-toe game. Specific tasks requested by the user will be entered and execution results will be presented through a GUI. An individual procedure is used to handle each task. For instance, routines are written to fetch and save data from/to a file. Students are also required to write code to handle common exceptions. Later in another experiment, database techniques will be introduced to provide students with an alternate means for data management. Visual Basic – Web Programming. The objective of this experiment is to provide tools for students to implement a three-tier application. To this end, students will learn to create a web application using Visual Basic.NET. The web application will run under MicrosoftÕs Internet Information Server (IIS). Eventually, the web application will have a number of tiers. The front tier consists of a set of web pages, interacting with remote users through a web browser. The middle tier links the remote user to data sources and plants. The back tier, in our case, is a Process Server, which will be discussed in detail in a later experiment. The web pages created will be decorated with graphics/image files such as logos and photos of the laboratory. From the pages, remote users should be able to log in to the site, fetch documents, such as laboratory instructions, and navigate to other locations with ease. In the homework assignment, students will move the math tutorial and tic-tac-toe applications to the web. Technically, the students have to divide each task into two parts: the GUI part and the
338
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344 Actuator
Physical Plant
Sensor
(Servo Motor)
(Robot)
(Encoder)
Controller 2 (Basic Stamp 2)
Transceiver 2 Transceiver 1 Controller 1 (Basic Stamp 1)
GUI (Windows Application)
Fig. 4. A diagram for RF remote control.
algorithm part. The GUI part will be similar to that of a window application, except that it is implemented using web components and will be delivered to a remote user through a browser in this case. The algorithm part, also similar to the one in the windows application, will be coded in a program that runs at the server side. Basic Stamp II. In this experiment, students will learn to program Basic Stamp microcontrollers. Basic Stamp is a single-board microcontroller that runs the Parallax PBASIC language interpreter. The developer stores all the code in an EEPROM,2 which can also be used for data storage. The PBASIC language has easy-to-use commands for basic I/O, such as turning devices on or off, interfacing with sensors, etc. More advanced commands let the Stamp interface with other integrated circuits, communicate with each other, and operate in networks. In this experiment, students will learn how to program Basic Stamp II, build a simple plant (i.e., a robotic cart and some parallel LEDs) and write a PBASIC program to control the plant. Control the Robot with Wireless Links. Once students are familiar with Basic Stamp microcontrollers, they are then ready to implement a simple control system. First, they will design a GUI as a Windows application in Visual Basic. They then write code to communicate the control commands and feedback signals via two Basic Stamp microcontrollers, both of which are interfaced with a pair of transceivers, as shown in Fig. 4. Through the experiment, students will interpret the control loop as follows: A user initiates a command by clicking a button on the GUI. The Windows program behind the GUI sends the command to Basic Stamp 1 through a serial communications port using a communication protocol (e.g., RS 232). Basic Stamp 1 talks with Basic Stamp 2 through a RF (radio frequency) link (the distance is up to 200 feet). Basic Stamp 2 interprets the command and sends a control signal to the physical plant. The feedback signal from the plant can be sent back to the user through the path in reverse order. Robot control system via the internet. Once the robot is properly controlled locally, the entire routine developed for the system will be ready to be moved to the Web. In this experiment, 2 EEPROM: electronically erasable programmable read-only memory, which can be programmed repeatedly and can keep data even after the power is shut off.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
339
students will control the robot remotely using web applications with transceivers and Basic Stamps. They are required to interface the robot with the Process Server through the transceivers and perform communications between the Process Server and the Web Server. They will also program the web camera to send image sequences of its environment to the remote user, providing visual feedback to the user. A block diagram of the control strategy is given in Fig. 5. A large portion of the diagram is identical to that of Fig. 4, except the Internet part. A user enters a command through a GUI at the client side. A web browser will send the command to the Web Server that hosts the web application. The web application creates a TCP client socket that communicates with the TCP server socket residing at the Process Server. The Process Server does the same job as the Windows application shown in Fig. 4. The remaining control actions are identical to those shown in Fig. 4. As previously mentioned, students can either take a final test or develop a project after completing the experiments. Those who choose to do a term project can either create an Internet-based control system similar to the one shown in Fig. 6 or develop a control system of their choice with the instructorÕs approval. The GUI for the servo control in Fig. 6 includes PD parameter settings and other functionality that students might want to add on. Moreover, the control plant is monitored by a web camera that is installed in the laboratory. A remote user selects the control type and sets control parameters through the GUI. The control command is then sent to the plant via the control path similar to that given in Fig. 5, in which case a data acquisition board from National Instruments is used to replace the Basic Stamps and wireless link. The controller, implemented in the
Actuator
Physical Plant
Sensor
(Led, Servos)
(Led, Robot)
(Led)
Controller2 (Basic Stamp2)
Transceiver2
Transceiver1 Controller1 (Basic Stamp1)
Server Socket (Process Server)
Client Socket (Web Server)
GUI (Browser, Cli ent side)
Fig. 5. A diagram for Internet-based control.
340
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Fig. 6. A typical graphic user interface.
Process Server, generates a control signal and sends it to the servo motor, which in turn moves the disk to the desired location in a closed loop. 4. Assessment and discussion In the summer of 2003, this course was offered for the first time to 20 junior/senior engineering students after about a year of preparation. The course was conducted in a six-week period, including classroom instruction and laboratory exercises. Students completed the first three laboratory exercises individually. Due mainly to the limited number of Basic Stamp microcontrollers, transceivers and robots, they completed the remaining three exercises and the optional term projects in teams (each team had two students.) At the end of the term, a laboratory test or a term project was conducted in the class to judge the achievements of the students in the following topics: 1. 2. 3. 4.
Create windows and web applications. Program Basic Stamp II microcontrollers. Interface PC with microcontrollers. Create windows and web applications to control devices through microcontrollers.
About 60% of the students chose to conduct the term project, and the rest did the laboratory test. Among the students who did term projects, a couple of teams produced excellent results. One team constructed an Internet-based soil moisture control system and another, an Internet-based temperature control system. Both projects involved sensing, Internet programming, and control. As to the laboratory test, students were asked to: (a) create a web application with a simple GUI to allow users to enter commands that, say, turn on and off a LED which is interfaced with a Basic Stamp; (b) send the user commands to the Basic Stamp through a serial port of a personal computer; and (c) carry out the user command using the Basic Stamp. The test seems simple, but it contains most of the components of an Internet-based control system. About half of the students completed the test without much difficulty. The remaining students had problems during various steps of the test. They were given hints to overcome hurdles, with some penalties. At the end, all of the students completed the test.
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
341
At the end of the class, a questionnaire was distributed to the 20 students. The answers to sample questions were compiled and are given below. The numbers in the brackets are the number of responses. Some students gave multiple answers. Question 1: What was your original intention of taking the class?
To control electrical devices via the Internet (5). To learn user-friendly computer languages such as Visual Basic (5). Need an elective (4). To learn how to program microcontrollers such as Basic Stamp (3). It sounds very interesting to topic (2). To learn computer interface (2). To have a class with practical application and hands-on experience (2).
Most students expressed a desire to explore and learn new things. A student stated that she would like to find out more about web-based control to see if she would be more interested in it for the field (work). There were a few students that just wanted to have an elective to fulfill graduation requirements. Question 2: Has the original objective been achieved after taking the class? All the students except one stated that their original objectives were met. Many indicated that this was an interesting class that opened up their minds. They also stated that the knowledge learned could be applied to other classes like Engineering Design I and II. A majority of the students pointed out, however, that six weeks were too short to cover the materials in depth and more time was needed. The student who answered no to the question also explained that he needed more time to digest the material. Question 3: What is your opinion on the method of alternating between lecturing and laboratory exercises? A vast majority of students favored the integration of lecturing with the laboratory exercises. They stated that the laboratory exercises provide much needed hands-on experience that reinforces the learning. Some students even suggested that the lectures should move to the laboratory. One student suggested that the material provided in the lecture should match more closely what was exercised in the laboratory. The only student who gave a negative reply to the question proposed to have the first half semester on the theory and the second half semester on the laboratory exercises. Question 4: What is your opinion on the laboratory equipment? Most of students thought the laboratory setup was good. Basic Stamp is a very user-friendly microcontroller, and Microsoft Visual Basic.NET is a very powerful, yet easy to learn, tool to create web applications. Some students complained that Basic Stamp educational boards and the robots used in the laboratory were not sturdy enough. Some stated that more computer stations were needed to allow everyone to conduct the exercises independently. Question 5: What is your opinion on the lecturing? Most of students gave positive responses to this question, though they have, in general, more favorable opinions on the laboratory exercises than the lectures. They like practical examples more than pure theoretical discussions. They prefer more simple demonstrations to show them step by step how things work. They want to also see more discussion on the equipment used in the laboratory.
342
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
Question 6: What is your opinion on the laboratory exercises? Students had, in general, very favorable opinions on these laboratory exercises. They stated that the way to arrange the laboratory exercises from simple to more complex ones was thoughtful. The fact that the last laboratory was an integration of the previous laboratories was especially well received. Some students were very excited about the success of controlling a robot through a wireless link over the Internet. One student said that creating a website to control something was very educational. Question 7: Any other comments? Sample answers to this question are given below:
This should be one of the first courses students are given. Overall good educational class. Great tool for my senior design (Engineering Design I and II) project. Very good class. One of the few classes I have taken with a practical application. Time went by too fast. Good class. Nice to see some more current technologies in (other) classes. Best wishes next year.
Since the number of students that took the survey was relatively small, the results cannot be considered conclusive. We feel, however, that the survey provided us with useful information that is being used to improve the course for subsequent offerings. The performance of the students that had taken the Internet-based Instrumentation and Control (IIC) course was tracked to see how their performance fared in the Capstone Engineering Design I and II (ED I and ED II) courses. Table 1 lists the grades of these (20) students in comparison with the entire class (about 90 students). The grade scale is 0–4, with 4 being the highest. A simple hypothesis test is conducted with the data. The null assumption is ‘‘there is no performance difference between the two groups of students’’. Since the sample size is relatively small, a t-test is used in this study. The level of significance a is selected to be 0.05. The corresponding right-tail critical value Za is 1.645. The standard deviation s for both ED I and ED II for the first group is approximately 0.5. The t-statistic in this case is x l t¼ s ; pffiffi n where the sample mean x is given in the 2nd column of Table 1, the population mean l, is approximated by the class mean given in the 3rd column, the sample standard deviation s is 0.5, and n is 20. The computed t statistic for each case is given in the last column of Table 1. These are much smaller than the critical value 1.645; therefore the null hypothesis is rejected. Based on the sample, there is sufficient evidence at the 5% significance level to conclude that the students who had taken Table 1 Performance comparison of the controlled group with their peers ED I grade average ED II grade average
Statistics of students who had taken IIC
Statistics of the entire class
t Statistic
3.58 3.65
3.25 3.29
2.87 3.13
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
343
Fig. 7. Pablo Sanchez, a member of the FAU team, with its robot (Boca Raton News, FL, 4/15/2004).
the Internet-based Instrumentation and Control course achieved better results in the Engineering Design II course. In a robot competition sponsored by the American Nuclear Society in Spring 2004, Florida Atlantic University (FAU)Õs teams came out on top. FAU sent four teams of undergraduate engineering majors to the University of Florida campus for the competition. The task was to pick up three items of ‘‘nuclear waste’’ that had been dropped on the floor in a ‘‘hot room.’’ Each team had to use a robot to scoop up the waste items and contain them in a least amount of time. The FAU students did a superb job in the competition and took the top two spots. Four of these students had taken the Internet-based Instrumentation and Control course the previous semester. Their success was due to a number of factors, one of which we feel was the hands-on experience in microcontrollers, windows programming, and sensing and control that they gained in the Internet-based Instrumentation and Control course (Fig. 7). 5. Conclusion The primary objective of this paper is to describe the development of a course for undergraduate engineering students who are interested in learning and applying Internet programming technology to tasks in Internet-based control. Based on the positive response of the students who have taken the course, it is our belief that the course is a viable elective engineering course that fits Electrical and Computer Engineering curricula very well. Future work includes refining the course, writing a textbook for this unique course, disseminating the results to other engineering colleges, and offering a similar course to high school students who are interested in pursuing engineering degrees.
Acknowledgements Efforts made by our graduate students, Cristian Popescu, Qingmei Li, and Hesong (Harry) Huang in preparing this course are greatly appreciated. The support provided by the National Science Foundation for the laboratory development component of the course is gratefully acknowledged.
344
H. Zhuang, S.D. Morgera / Computers & Education 49 (2007) 330–344
References ABET Engineering Criteria. (2000). Accreditation Board for Engineering and Technology (ABET). Butner, S. E., & Ghodoussi, M. (2003). Transforming a Surgical Robot for Human Telesurgery. IEEE Journal of Robotics and Automation, 19(5), 818–824. Dutta-Roy, A. (1999). Networks for Homes. IEEE Spectrum, 26–39. Furuya, M., Kato, H., & Sekozawa, T. (2000). Secure Web-based monitoring and control system. In Proceedings of the 26th annual conference on IEEE industrial electronics society, Vol. 4, pp. 2443–2448. Guangcheng, M., Fei, T., Changhong, W., & Yufeng, W. (2003). Web-based control service for self-organization of IDCS. In Proceedings of the IEEE conference on control applications, Vol. 1, pp. 1198–1202. Jeon, J. M., Kim, D. W., Kim, H. S., Cho, Y. J., & Lee, B. H. (2001). An analysis of network-based control system using CAN protocol. Proceedings of the IEEE international conference on robotics & automation, Vol. 4, pp. 3577– 3581. Kato, H., Furuya, M., Tamano-Mori, M., Kaneko, S., & Nakano, T. (2001). Risk analysis and secure protocol design for WWW-based remote control with operation – privilege management. In Proceedings of the IEEE international conference on systems, man & cybernetics, Vol. 2, pp. 1107–1112. Lin, A., & Broberg, H. L. (2002). Internet-based monitoring and controls for HVAC applications. IEEE Industry Applications Magazine, 8(1), 49–54. Overstreet, J. W., & Tzes, A. (1999). Internet-based client/server virtual instrument designs for real-time remote-access control engineering laboratory. In Proceedings of the American control conference, Vol. 2, pp. 1472–1476. Park, J. W., & Lee, J. M. (2001). Transmission modeling and simulation for internet-based control. In Proceedings of the 27th annual conference on IEEE industrial electronics society, Vol. 1, pp. 165–169. Ramakrishnan, V., Zhuang, Y., Hu, S. Y., Chen, J. P., Ko, C. C., Chen, B. M., et al. (2000). Development of a webbased control experiment for a coupled tank apparatus. In Proceedings of the American control conference, Vol. 6, pp. 4409–4413. Tan, K. K., Lee, T. H., & Soh, C. Y. (2002). Internet-based Monitoring of Distributed Control Systems-An undergraduate Experiment. IEEE Transactions on Education, 45(2), 128–134. Wireless Networks and Web-based Education, a Special Issue. IEEE Control Systems Magazine 23(2), April 2003. Yang, S. H., Tan, L. S., & Chen, X. (2002). Requirements specification and architecture design for Internet-based control systems. In Proceedings of the computer software & applications conference, pp. 75–80. Zhang, J., Chen, J., Ko, C. C., Chen, B. M., & Ge, S. S. (2001). A web-based laboratory on control of a two degree of freedom helicopter. In Proceedings of the IEEE decision & control conference, Vol. 3, pp. 2821–2826. Zimmer, T., Kadionik, P., & Danto, Y. (1997). A World-Wide-Web based instrumentation pool real testing in a virtual world. In Proceedings of the 6th annual conference on IEEE industrial electronics society, Vol. 4, pp. 2443–2448.