Written late 1990's-ish, updated August 2003

Designing an Electronic Classroom

Every year or so I come across someone excitedly talking about how technology is going to enable great advances in the educational process, in particular in the area of distance learning.

Invariably, video recording and transmission of lectures comes up as the number one important idea. The idea is you can video the lectures and transmit this video live to remote sites, where students can watch, and even interact. The lectures can be recorded so that students can review the material later on (for missed classes, or to go over areas that were confusing the first time). The instructors can also review the material each semester to improve their lecture style and content.

This is fine as far as it goes, but as they say, the devil is in the details. The particular place the devil lives here is in the video. There are some major problems with videoing a lecture in this way.

The biggest problem is the resolution of the video camera. In a conventional classroom setting, with the instructor using a blackboard (or two or three), you'd want to video everything on the blackboard. That is (usually) more important than to show what the instructor looks like, or is doing. But if you point a camera at a big blackboard area, you won't be able to read any of it from the output of a video camera, because the text will be too small on the TV to be seen (or the camera won't cover all of the blackboard area, and some of the text won't even be in view).

This necessitates having a human operator for the video camera to pan and zoom as the instructor works, which means either additional expense to pay an operator, or distracting the students by having them volunteer to run the camera, reducing both the quality of the classroom time, and the quality of the video.

Another major problem comes if you want to record these lectures during the semester or school year so that students can go back and review them. This would be a substantial tape library. To provide remote access to that library would involve a very expensive tape robot system, or a very expensive tape librarian.

Storing this digitally would also be prohibitively expensive due to the enormous size of the video data files. A one hour class, meeting five times a week for 20 weeks would need approximately 100 Gigabytes of space - multiply that by the number of courses you offer. If you think it is important to record each section of each course, multiply again.

A better way

What you really want to capture is what the instructor says, and what goes on the blackboard. Video is poorly suited to do this. What is needed is to replace the blackboard with something new. There have been some interesting experiments here, some even worth pursuing. I'm going to describe my own ideas, which can be accomplished (relatively) cheaply, with off-the-shelf items.

Instead of a blackboard, use one or two video projectors. Instead of a normal podium, the instructor has a computer with one or two monitors, a light pen or writing tablet, and a keyboard and mouse. Also, a microphone for recording what the instructor says. The instructor would use some sort of whiteboard application.

The problem of resolution is solved: computers have much higher resolution than conventional video. Also, whatever the instructor does has to be visible on the classroom displays, and therefore will be visible remotely or on the recordings.

Recording size is also solved - recording audio takes up much less spess than audio+video - the example given above that required 100 Gb of disk space for video would only need about 1.5 Gb for the audio. As for the activity on the whiteboard, this can be recorded in even less space than the audio because you don't have to record each frame, only the activity.

(Suppose the instructor draws a figure, and takes a five seconds to do this. Recording frames, this would take 150 frames, with each frame requiring at least 100K for compressed computer-resolution images - this would be 15 Meg. The same data based on activity, would store a series of pen positions and times, which would take about 12 bytes per sample. If you stuck with video frame rates, you'd need 150 samples, or 1800 bytes. A semester of such data would require about 130 Meg.)

There'd be some nice bonuses to this sort of approach that would also improve the classroom:

The only missing piece is the software. There may be software for teleconferencing that has many of the desired features. But for complete success, such a project would need it's own dedicated software. Fortunately, universities have no lack of bright eager software developers who will do the work cheaply or even for free. Here's what you'd need to start: An application with such basic functionality could probably be done by a single programmer with the right skill set in less than a month. Later you could get fancy, give the lecturers powerful tools for preparing lectures, the ability to prepare animations, even to integrate the occasional short video segment. The playback applications could gain the ability to search, to smoothly rewind, etc.

I've always believed in the use of Technology in education. Unfortunately, it always seems to be misapplied. The system I've described above would apply technology to legitimately improve a current tool of teaching, the blackboard, rather than cluttering the class with new, poorly integrated technology that hinders rather than helps.


August 2003 Update - ScreenWatch

The CEO of OPTX, the makers of ScreenWatch, wrote to inform me of their product, probably with the hopes that I would take a look at it and comment here, and that's exactly what I'm doing. Note that I have never used their product, and so I'm only commenting on their concepts.

ScreenWatch takes a somewhat different approach than I imagined. Rather than providing a blackboard application, under which you can do any presentation you need, ScreenWatch provides a recording of what is on the monitor, and the ability to mark it up. This means that it doesn't have to reinvent powerpoint - you can just use powerpoint, or your favorite spreadsheet, or any other program that you want to integrate into your presentation.

This is very powerful - it is easier for the lecturer to use the tools they already know, and reduces the complexity of ScreenWatch considerably. However, I have to guess that this does not result in the tightly integrated solution that I think would be ideal.

The recording is supposedly done in a fairly efficient fashion, recording only screen changes, and not screen frames. This would not be quite as efficient as my vision, but should certainly create manageably sized recordings if they implemented the recording as they've claimed.

They claim good speed of playback over a dialup modem, which is reassuring. It isn't clear if they offer live broadcasting of the lectures too.

I've never used the product, I'm not exactly endorsing it, but I will say that it appears to be a step in the right direction, and is worth investigating.


My Home Professional Home TomSaraZac Home
Email Work Email