This paper describes an Internet-based course on Beam Physics being offered by the Department of Physics (http://www.pa.msu.edu) at Michigan State University (http://www.msu.edu) in the spring term of 1997 (http://www.beamtheory.nscl.msu.edu/lectures/phy861). This course is part of the MSU Virtual University program (http://www.vu.msu.edu) and has about 100 registered participants at approximately 25 different sites all over the world. For the purposes of this course we are using ISDN-based (PictureTel) and Internet-based (CU-SeeMe) videoconferencing tools, Internet-based transmission of audio and video recordings of the lectures (RealAudio, RealVideo), an interactive Internet-based homework system with on-line grading (CAPA, http://www.pa.msu.edu/educ/CAPA/), and we provide the participants with downloadable lecture notes in a variety of formats.
Beam Physics is one of the more recent subfields of physics, and is connected to the understanding and development of particle accelerators, some of which represent the biggest and most expensive scientific instruments currently in existence. Because accelerator laboratories are usually not directly connected to university environments, the proper training of beam physicists at these sites often does not happen as naturally as in other fields, and a significant amount of instruction is provided by the US Particle Accelerator School (http://fnalpubs.fnal.gov/uspas/welcome.html) via biannual two-week block courses at various locations in the USA as well as some other similar institutions. The availability of video conferencing and other Internet-based tools offers students and employees an option of increasing or refreshing their knowledge of Beam Physics. This approach provides an efficient and inexpensive mechanism to learn in a systematic fashion and offers the opportunity to earn university credit without leaving the workplace.
In order to facilitate access to support and tutoring, at the major sites there are one or two experts serving as local contacts to help with organizational issues, homework, etc. To increase the quality of the course and the contact with the participants, we are offering office hours using the same technology as for the course itself as well as a web-based news bulletin that keeps the participants informed on course developments. All information relevant to the course including schedule, technical setup, lecture notes, lecture audio files, as well as the current class list are available from the course home page (http://www.beamtheory.nscl.msu.edu/lectures/phy861/index.htm). For the future we are considering to offer the course on a stand-alone basis in which a student can participate at his own pace by viewing existing lectures and solving the automated homework problems. It is also planned to extend the concept to other courses, and to utilize CD-ROM as an additional format of dissemination.
Introduction
Beam Physics became an important subfield of physics which incorporates many important practical applications and challenging theoretical aspects. It is connected to the probing of the fundamental properties of nature and the search for new physics through high energy accelerator experiments, which represents the largest scientific experiments, to elucidating the structure of huge biological molecules trough mass spectrometers, to visualize tiny surface details trough electron microscopes, to fabricate computer chips trough micro beam litography, to build CRT's for TV sets, to separate isotopes, to measure exotic nuclei, and a variety of other techniques.
New theoretical methods developed for beam physics are at the forefront of physics, and they not just facilitate the understanding of current scientific instruments and provide solutions for future ones, but they are of academic interest by themselves, and in many cases their applicability goes far beyond the domain of beam physics. These are some of the reasons why a few years ago the Division of Physics of Beams was created within the American Physical Society. Nevertheless, there are many inconveniences and obstacles related to Beam Physics education. Due to the nature of this field, the highly trained instructors and specialists are spread over universities and major research laboratories. In the United States, only a few universities provide Beam Physics curricula, and one of them is Michigan State University.
From the above reasoning, the necessity to organize such a tele-course became clear; the goal is to provides large scale access to Beam Physics instruction, even internationally. Beside that, there are many further advantages: the convenience of taking such a course without leaving the workplace or school: reduced cost, flexible scheduling, the ability to gather recognized specialists to give guest lectures on their topic of expertise, etc. This paper describes what has been accomplished in this direction, and we think it provides the first step toward the future's virtual classroom.
The following table gives an overview of the participating sites, locations and number of participants per site.
| Site Name | Country | Number of participants |
| Argonne National Lab. | USA | 21 |
| Beijing University | CHINA | 1 |
| Brookhaven National Lab. | USA | 8 |
| Calcutta University | INDIA | 2 |
| T. Jefferson National Lab. | USA | 11 |
| Cornell University | USA | 1 |
| DESY | GERMANY | 1 |
| Dubna Laboratory | RUSSIA | 2 |
| Fermi National Accel. Lab. | USA | 7 |
| Kansas State University | USA | 2 |
| KVI | NETHERLAND | 4 |
| Los Alamos National Lab. | USA | 1 |
| LawrenceBerkeley Nat. Lab. | USA | 6 |
| Lawrence Livermore Nat. Lab. | USA | 2 |
| Mississippi State University | USA | 1 |
| Michigan State University | USA | 11 |
| Sandia National Lab. | USA | 1 |
| Stanford Linear Accel. Center | USA | 1 |
| St. Petersburg State University | RUSSIA | 5 |
| Stony Brook Laboratory | USA | 1 |
| TRIUMF | CANADA | 3 |
| University of Chicago | USA | 1 |
| Univ. Of Illinois, Chicago | USA | 1 |
| Universite Laval | CANADA | 1 |
| University of Helsinki | FINLAND | 1 |
Table 1: Participating Sites
The subsequent sections will give some information about the importance of local contact persons, the technical aspects, used equipment and the Internet based homework assignments we have used. Finally we discuss the shortcomings and difficulties we have encountered and summarize our conclusions and future plans.
Technical Details
The method of attending the lectures allows a classification of the participants of our course in 3 major groups, with a subdivision of one of them in two subgroups.
P1 - PictureTel User
A - Online User
B - Off-line User
P2 - CU-SeeMe User
P3 - Real Audio User
The chart at the end of the paragraph gives a schematic overview of the equipment used and the flow of information to the participants for the various groups.
P1 - PictureTel
The major lab sites in the US are equipped with ISDN-based teleconferencing tools like PictureTel [PT]. These sites do use this equipment to participate in the lectures. Based on ISDN, the integrated digital data service, it offers higher bandwidths than ordinary phone lines and wide availability. While these solutions by themselves offer only point-to-point connections, Lawrence Livermore National Laboratory in California offers a service that allows more than two sites to connect to one video conference [VCS]. We are using this service (bridge) to serve the PictureTel based participants. This service is provided to institutions supported by the US Department of Energy for free; it is also available commercially at a rather affordable rate of about $40/hour. But the equipment for the system lies in a price range of about $20,000 up to $100,000, which naturally restricts the users of this system to participants located at the major lab sites in the US and Europe. While this system in principle has no restrictions on the locations (in fact we successfully transmitted lectures from Hamburg, Germany), all but the Hamburg site are located in the US.
Since the participants in this system actually see the other participants on a regular TV screen and the video signal is highly compressed, special arrangements have to be made in order to transmit a readable picture of the lecture notes. A designated document camera is used and the text has to be slightly bigger than usual.
Due to the time shift in the US and our lecture schedule (we start at 9:45AM EST), the lecture takes place rather early in the morning at the West Coast. The participants at these places record the lecture with a normal VCR and do watch the lecture later off-line (Group P1B). The members of Group P1A do attend the lectures live and online.
P2 - CU-SeeMe
The second group of our participants are the CU-SeeMe users. CU-SeeMe is a videoconferencing technology developed at Cornell University that uses regular TCP/IP to transmit highly compressed video and audio data over the Internet [CU]. The major advantage of this technology is that the necessary soft- and hardware is free or rather inexpensive. What all the participants of this group do actually need is a personal computer (either Windows based or a Mac), a working network connection (fast modem connections are in fact sufficient) and a sound card. Altogether these equipment needs can almost be considered standard for modern personal computers.
In addition, the participants do need the CU-SeeMe client software. There are two choices to obtain the software: it can either be purchased from a commercial vendor, or one can use the freeware client from Cornell University. Although the commercial product has some enhancements over the free version, we do not use it in order to maintain compatibility with the participants that do use the free version. Even the commercial product is at a price of $80 and hence rather affordable. On the server side a Windows-based PC with a frame grabber card is needed, at a cost of currently in the range of $250. It is not necessary to use commercial software in order to transmit the signals. In order to operate the CU-SeeMe part of the video conference, the participants log in to the deflector, which - as its name suggests- mirrors the incoming videoconference signal to the participants.
P3 - Real Audio
In order to provide participants that can not attend the lecture live with as much information as possible, we make an audio recording of the lectures available on the Internet. These audio files are encoded from the video tapes that we have recorded during the lecture. The file format is the Internet standard Real Audio from Progressive Networks [RA]. The participants can download these files a few hours after the lecture and listen to them while they follow the lecture notes that are available on the web.
This method of participating is especially suitable for participants who cannot afford the PictureTel equipment, have very slow Internet connections to US sites, cannot take the lecture due to time shift problems, or who just missed the class for whatever reason. Due to recent developments by Progressive Networks towards a high-compression video standard for off-line viewing, we hope that in the near future it will be possible to provide the off-line participants not only with audio data but also video recordings of the lecture. However, for users with low bandwidth connections, the resulting files may become too large, considering that the audio files along require about eight megabytes per lecture.
It became clear early on that some participants would not feel comfortable with an Internet-based course that lacks the personal contact with the main instructor. Therefore it was of prime importance to us to establish close connections to qualified local contact persons that can help the students with questions and can provide personal contact. In some places these instructors even formed small groups with their students that were working together through the course material in local lectures. To provide further assistance for the participants, we established local office hours. During a fixed time slot we were offering the students the possibility to reach us by phone, fax, electronic mail and CU-SeeMe to ask questions regarding the lectures and the homework problems. We think that the general aspect of accessibility of the main instructor and the availability of help closeby cannot be stressed enough in the preparation of any distance education course.
Homework
For homework assignments, we are using CAPA [Ca94,95,97], which is a software tool to implement a Computer-Assisted Personalized Approach for homework assignments, quizzes, and examinations. It was developed in such a way that it provides each student with a personalized assignment or examination with both quantitative and conceptual qualitative questions. With CAPA, an instructor can create problem sets which include pictures, graphics, tables, etc., with variables that can be randomized and modified for each student. Students input the solutions via a standard web browser, are given instant feedback and relevant hints, and may correct errors without penalty prior to the assignment due date. The system records the students' participation and performance in assignments, quizzes and examinations; and records are available on-line to both the instructor and the individual student.
CAPA was developed through a collaborative effort of the Physics-Astronomy, Computer Science and Chemistry Departments at Michigan State University, and the current version 4.5 became available April 15, 1997. However, more advanced features are needed in order to assign homework problems suitable to graduate students. Currently, the basic types of problems include numerical, multiple choice, matching, and true-false types, but there are no provisions for analytical derivations. At present, these type of problems can be hand- graded, and then the earned points added manually to the Grader module of CAPA. However, even in this case, the instructor saves some time compared to the completely hand-graded style. Of course, CAPA becomes clearly time saving for big classes. Future developments will provide support for other type of problems as well.
Shortcomings of Current Technology
While we were preparing this course we were confronted with the need for evaluating the currently available mechanisms to publish scientific text on the WWW. One major aspect was the need to have a tool that allows the transformation of texts that are written in LaTeX, the main text editing software supporting complicated mathematical expressions, to HTML documents. Since most scientific texts containing extensive mathematics are written in LaTeX format, it seemed natural to us that there should be a way to publish these documents on the web. Two major aspects have been important for us. First of all the transformation should be as easy and as compatible with any computer platform as possible, and second of all the result should be esthetically pleasing.
While there is a tool that transforms LaTeX input files to HTML documents (LaTeX to HTML) the results are unfortunately often far from being pleasing, since every single equation is transformed into a separate GIF file. The resulting documents loose nearly all the nice formatting typical of LaTeX, and due to the literally hundreds of GIF files the loading times of the resulting HTML documents are unacceptable. Besides this tool there are some other solutions for LaTeX publishing on the web (TeX Explorer from IBM - a Netscape plug-in for Windows [TE], Scientific Notebook - a proprietary browser from TCI Soft [TCI]). But all of them have the shortcoming that the software solutions are restricted to certain browsers and/or certain platforms. Since we had to maintain compatibility to all possible platforms, these solutions ruled themselves out.
While there are also Java applications that can display mathematical equations in a very nive way, these solutions have the major shortcoming not to understand LaTeX input. Since the current HTML specifications do not contain any commands to support mathematical equations (in fact a draft for this feature has been removed from the current version 3.2), we are actually not very optimistic about the future development in this direction. Moreover the suspended draft for mathematical expressions in HTML 3.2 did not claim any compatibility with LaTeX. Therefore it seems to us that the publishing in PostScript format with all its shortcomings will remain the standard for publishing scientific texts containing mathematics in the near future. The only promising approach towards this problem seems to be the future development of the Portable Document Format (PDF) by Adobe [PDF].
Another major problem we had to deal with was the bandwidth restriction of today's Internet backbones. The Internet in its current stage is not as fast and stable as one would really need for continuous audio and video transmission in high quality. Not only that participants from abroad had difficulties in using the CU-SeeMe based conferencing tools, but also students at sites in the US had problems with real time transmission of data over the Internet (our experiments showed that there are packet loss rates of up to 90% during the day on connections within the US). Even the normal download of big lecture notes and audio files from our server (which was not the bottleneck) turned out to be a problem.
It is expected that this problem will be alleviated substantially with the introduction of Internet 2, at least for those sites that have easy access to one of its hubs. Together with the rapid increase of Internet-based activities, it appears that it is only a question of time to make a continuous data flow over the Internet (at least within the US) possible to allow real time video conferencing over the Internet.
Conclusion and Future Developments
Overall we rate the Internet- and video conferencing-based course in Beam Theory we have given in the spring term of 1997 at Michigan State University as a full success. Not only have we reached a wide audience, but we also gained experience that we will use in future projects on distance education.
Since all the material is available in a web browser readable format, it is a natural extension to produce a CD- ROM out of the course material. This would allow students to take the course as it fits their needs and independent of the curriculum at Michigan State University that offers this course only in odd-numbered years. Given the interactive homework approach via CAPA, it is even possible to award credit to each individual who has at his own pace completed a full set of CAPA problems. This CD-ROM could be viewed with any web browser and could even contain the necessary additional programs (like the Real Audio player and GhostView).
Last but not least we are already thinking about the next Internet-based course in physics that will come in the near future. Using the experience that we have gained now, we are looking forward to improve our methods and organization with regard to distance educational projects.
References
[Ca94] E. Kashy, D.J. Morrissey and Y. Tsai, CAPA - System Description and User's Manual, MSUCL-925, January 1994.
[Ca95] E. Kashy, D.J. Morrissey, Y. Tsai and S. L. Wolfe, CAPA - A Versatile Tool for Science Education, MSUCL-971, September 1995.
[Ca97] http://www.pa.msu.edu/educ/CAPA/
[CU] http://cu-seeme.cornell.edu/
[PDF] http://www.adobe.com/supportservice/devrelations/PDFS/TN/PDFSPEC.PDF
[PT] http://www.picturetel.com/
[RA] http://www.realaudio.com/
[TCI] http://scinotebook.tcisoft.com/scinotebook/default.htm
[TE] http://www.ics.raleigh.ibm.com/ics/techexp.htm
[VCS] http://vcs.es.net/vcs
Martin Berz (berz@pa.msu.edu) Bela Erdelyi (erdelyib@pilot.msu.edu) Jens Hoefkens (hoefkens@pilot.msu.edu) Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University, East Lansing MI 48824, USA