SHARING ONLINE LABORATORIES AND THEIR COMPONENTS - A new learning experience

SHARING ONLINE LABORATORIES AND THEIR COMPONENTSA new learning experienceKJELL O. JEPPSON, PER LUNDGREN, JESUS A. DEL ALAMO*, JAMES L.HARDISON*, DAVID ZYCH*Chalmers University of Technology, MC2, SE-41296 G\uf6teborg,Sweden,*Massachussetts Institute of Technology, Cambridge, MA, USA1. INTRODUCTIONIn this paper we report on the use of the online MIT Weblab system [1] forcharacterization of semiconductor devices in three qualitatively rather diversemicroelectronic device courses offered by Chalmers University of Technology,including junior undergraduate courses as well as extension courses. In particularwe will focus on the learning situation and the impact of class size. Since thelaboratory equipment is available online 24 hours-a-day every day during thecourse, new opportunities for integrating laboratories into the learning processhave become available. In particular, we will discuss the role of assignmentformulation to support this new learning situation.In this paper we will describe our experiences from using the MIT onlinelaboratory to shift student focus from instrument handling to data analysis,parameter extraction, and model fitting. This can be done through rather open labassignments where the students themselves can organize the details of theirspecific task within the context of the overall objective of the laboratory exercise.2. ABOUT WEBLABIn a topic like microelectronic device physics the student learning experience canbe substantially enhanced by hands-on characterization of diodes and transistors.However, for a variety of practical and economic reasons universities have found itmore and more difficult to include such a laboratory component. A remote laboratoryavailable over the internet solves many of these concerns while largelypreserving, or even enhancing, the educational experience. Online remote laboratoriesnot only offer the possibility to perform traditional laboratory exercises in amore cost effective way, but they also make available to students more advancedinstruments than have traditionally been affordable. Many institutions in differentfields have explored this concept of an online laboratory. One such joint Europeanremote laboratory network is presently being developed within the EU Socrates/Minerva framework [2].Over the last few years, MIT has been experimenting with a system called theMIT Microelectronics WebLab. This system allows microelectronic devicecharacterization through the world wide web. Through WebLab, students can takecurrent-voltage measurements on transistors and other microelectronics devices inreal time from anywhere and at any time. The basic architecture of the system andits use in a variety of educational settings was reported in [3].The user interface for WebLab is a Java applet which duplicates the essentialfunctionality of the analyzers console, see Fig. 1, allowing the user to set up ameasurement for one of the devices that is currently connected to the system (thenecessary information about these devices is provided by the server when theapplet loads). When the user is ready to execute a measurement, the applet sendsthe measurement specifications to the server. More details of the WebLab systemand its graphical interface are given elsewhere [4].3. SHORT DESCRIPTION OF CHALMERS COURSES AND MISSION TASKThe WebLab has so far been used remotely in three different courses offered byChalmers University. Following two small test runs (one in an elective graduatecourse (eight users) and one in a extension course offered to professionals workingin local industry (six users), WebLab was employed in a large compulsoryjunior undergraduate course with about 330 students during the spring of 2003.In all courses students were given a clear objective of the laboratory task andwhat was expected of them. A simple instruction was given that advancedtechnology transistors of four different types were available through MIT WebLab.Examination of the lab assignment in the undergraduate course was performedthrough group meetings where an examiner directed individual questions to the labgroup members who were to respond with the help of a whiteboard. Individualcredits were rewarded to the group members according to performance in this oralexamination. The communication between MIT WebLab administration andChalmers course management was conducted by e-mail and for the two smallcourses the planning could be settled with some ten mails and replies.4. EVALUATIONThe online laboratory experiments were evaluated through detailed discussionsFig. 1. Screen-shot of WebLab graphical interface: the main window.with students in the graduate courses and through written review questionnaireshanded in by students in the undergraduate course. The overall impression on theuse of online laboratories among engineering program students was generally verypositive. A summary of the evaluation regarding system access and stability, userfriendliness, and educational value is shown in Fig. 2.051015202530354045MiserablePoorOKHighOutstandingResponsesA cce ssib ility and stability Use r friendliness Educa tional valueFig. 2. Summary outcome of evaluation questionnaire.Students appreciated most that they could decide themselves when to do thelaboratory exercises. This was perceived as less stressful [than traditional eveninglaboratory classes]. They also appreciated the graphical interface, becauseyou could see the graphs clearly and it was excellent when analysing data. Themost severe drawbacks were the system instabilities. The system was instablewhen many students were logged on simultaneously: First you could not accessthe home-page at all, and then once you managed to log in it kept on crashing.5. OUTCOME: THE CRITICAL DIFFERENCE OF A LARGE STUDENT CLASSThe use of WebLab in the undergraduate course at Chalmers was the largestand most ambitious deployment of WebLab to date. On Feb 25, 2003, there were134 characterization experiments executed in a single hour (on average thatmeans a job every 27 seconds). An experiment this scale was bound to result inthe identification of new problems that had never been seen before when operatingat lighter loads. Two kinds of problems were encountered. First, there was ahandful of system blackouts during which WebLab was unavailable for measurements.Second, the system returned an error message in response to a validexperimental setup. The origin of both types of errors was identified and corrected.6. DISCUSSION AND CONCLUSIONIn our experiment with online laboratories we had an explicit purpose to get awayfrom traditional closed-form laboratories. In that type of laboratory students devotemost of their time to handling of the instruments to collect device data following astep-by-step instruction manual, frequently asking the teaching assistant for help tofind a short-cut to the next step. In an online computer-based laboratory, instrumenthandling can be minimized through the WebLab graphical user interface.Thereby, the student focus can be shifted to data analysis, parameter extraction,and model fitting. In essence, online laboratories enables the students to take amore active role in defining the scope of the assignment they can do measurementswhen they feel ready for them and re-do them when and if need arises [5].However, new opportunities also mean new challenges, for instance when itcomes to finding a text book to support the new learning process. Traditionally,most text books do not contain detailed experimental sections. One exception isthe book by Pierret [6] that contains an excellent description of experimental diodesetups, but it leaves the readers on their own when it comes to transistorcharacterization. One wonders who will become the first text book author to includea description of the transistor parameter analyzer in their book?Some problems encountered in the course of this experiment had a morenegative impact on the overall experience of the undergraduate students atChalmers University when compared to MIT students using WebLab in MITcourses. There are two reasons for this. First, at Chalmers students worked ingroups, while at MIT students assignments were of an individual nature. This isrelevant because at Chalmers, students had to make an appointment to worktogether on their lab assignments at a specific time. If the system was not availableor if the system did not operate properly at that very time, students were forced toreschedule leading to frustration and possible project delay. In an individualassignment, a student has a lot more flexibility to schedule their work and theconsequences of system malfunction are much less severe.The second reason for the negative impact of systems problems in the overalleducational experience of the Chalmers students is the time zone differencebetween Chalmers and MIT. As a consequence, several hours could pass betweenthe occurrence of a system problem and its satisfactory resolution, this evenif the problem was of a trivial nature and its solution would only take a fewseconds.7. REFERENCES[1] J. A. del Alamo, J. A., L. Brooks, C. McLean, J. Hardison, G. Mishuris, V. Chang and L. Hui,The MIT Microelectronics WebLab: a Web-Enabled Remote Laboratory for Microelectronic DeviceCharacterization, World Congress on Networked Learning, Berlin (DE), 2002[2] R. Cabello et al, eMerge: An European educational network for dissemination of onlinelaboratory experiments, ICEE, Valencia (ES), 2003[3] J. Henry, Running Laboratory Experiments via the World Wide Web, ASEE Conference, 1996[4] J. A. del Alamo, V. Chang, J. Hardison, D. Zych, and L. Hui, An Online Microelectronics DeviceCharacterization Laboratory with a Circuit-like User Interface, ICEE, Valencia (ES), 2003[5] A. S\uf6derlund, K. O. Jeppson, F Ingvarson, and P Lundgren, The Remote Laboratory A NewComplement in Engineering Education, ICEE 2002, Manchester (UK), 2002.[6] R. F. Pierret, Semiconductor Device Fundamentals, Addison-Wesley, 199