|
Engineering standards are information rich documents which outline some prescribed practice. They are written in precise language, often referencing equations, tables, graphs or similar aids as well as prior standards. This is in order to ensure that the information presented is interpreted in just one way by all practitioners. As it is the detail that is ultimately of importance, the format of a standard is that each of the topics discussed is examined at length. This, though makes a standard a difficult document to browse. It is vitally important in engineering education that students are acquainted with standards. However, at most points in such an education program, what is of interest is the salient features of the standard, not the detail, and that demands a good browsing capability. This paper demonstrates how a multimedia approach may allow the information of a standard to be presented precisely, yet at the same time offer a browsing function of genuine value in educating the reader or otherwise broadly informing them. A prototype package has been developed for international computer bus standards. These are used to interconnect modules in many personal computers, workstations and industrial computer systems. The paper discusses its features and highlights the advantages of this approach.
There are many types of standards. In manufacturing, for example, there are standards relating to how to test raw materials and the manufactured products, on assessing the quality of the manufacturing process, on management of the manufacturing process to ensure the health and safety of workers and on the equipment actually used. Equally, there are many bodies which set standards ranging from local to national to international, and from government to private to professional institution. However, regardless of their creators and to what they are directed, there are two common points to almost all standards. One is that they are expected to be complete documents and so are very detailed covering all possible situations the user of the standard may encounter. The other is that they are written in very precise language designed to eliminate all possible ambiguity.
Eliminating misunderstandings and ambiguities is very difficult to achieve when a natural language is used. Partly as a consequence, standards now tend to be written in a very uniform style. The text can often be described as the verbal equivalent of equations in mathematical logic. Thus the grammar tends to be rather rigid. The text follows a single logical thread with each section as complete as possible which means repetition is quite common. However, the frequent need to refer to other sections of the standard sees all sections and equations numbered and referencing made via those numbers. The need for precise communication also sees much use made of graphs and tables.
These essential objectives of ensuring no ambiguity or misunderstandings can exist in what are very information rich documents creates three problems. One is that a standard is a very difficult document to read so there is a real possibility of omissions by the reader. Another is that this structure is not well oriented to browsing. This is a serious problem, as browsing is exactly what most readings of a standard are about. That is to say, the engineer merely wishes to confirm his or her memory of the issue against the standard. Finally, these properties make standards difficult to use in an educational context. They are written to aid the experienced practitioner in their duties and make no allowances for unfamiliarity on the reader's part. Clearly, though, standards should and must form a vital part of any engineering student's education.
The issue here as in so many other situations is that an information resource needs to be accessed in a way not very compatible with traditional printed forms. A means is needed of accessing this information in linear fashion, but there should also be non-linear modes for the particular needs of students and others. Graphs and equations do offer communication precision, but they are not alone in this. Many standards describe strict sequences of actions and these are often very well suited to animation. Thus the need is for different means of accessing this information in a way more compatible with human use, and this suggests using some of the accepted multimedia paradigms. This paper examines how engineering standards may be put into a multimedia form and reports on the success achieved with a particular application based on computer bus standards.
Modules are circuit boards usually with a male or female connector along one edge. Those boards are then plugged into another circuit board equipped with the opposite connectors. The number of these connectors present defines the number of slots in the structure This board can be of two types:
A computer bus standard essentially relates to how some given electrical connector is used. In addition, though, it usually specifies the size of module circuit boards, the maximum number of modules in a configuration, the maximum power drain of a system and similar details. The core of the standard has three key sections:
The signal lines in a bus are typically divided into three sections:
The IBM personal computer (PC) was initially released with a motherboard bus termed the XT bus. This was a very simple 8-bit bus intended to add basic I/O capabilities to the PC, Particularly with respect to printers and disk drives. The motherboard had a connector, but the modules themselves merely had printed circuit pins to contain costs. It was a very slow bus. Further, there was no standard per se, merely a description in an IBM Technical manual. IBM recognised its limitations and subsequently released machines with the Advanced Technology or AT bus. This bus extended the XT by adding some additional signals and allowing 16 bit data transfers. Physically, it simply added another connector to the motherboard to extend the functionality.
Within a few years, the XT/AT bus was beginning to become a significant limitation on PC performance and work commenced on defining a new bus. IBM evolved the Micro Channel Architecture (MCA) bus which allowed 32 bit data transfers at high speed. Now when the PC was first released, the major manufacturer was naturally IBM. However, as the IBM PC had an open architecture, a series of clone manufacturers commenced operations and began to take significant market share. IBM believed the MCA bus would be the industry standard and, in order to recoup past losses, proposed a licensing scheme which included a retrospective component based on the licensee's production of clones. The clone manufacturer's were incensed and, as they now comprised most of the market share, set out to create their own competing standard. They referred to the XT/AT standard as the Industry Standard Architecture (ISA) bus; a name which has remained. Their new standard was termed the Extended Industry Standard Architecture or EISA. Unlike MCA, this allowed older XT or AT cards to be used in a new system.
As has been shown in many situations, competing standards usually see customers make conservative choices. Even today, very few machines have either an EISA or MCA capability as most purchasers decided to remain with ISA. However, the need for high performance become critical. In particular, there was a number of difficulties encountered in achieving high performance graphics capabilities in PCs. In 1992, therefore, two proposals were put to industry for an entirely new bus termed a mezzanine bus. These would not replace ISA or MCA or EISA, but rather connect straight to the CPU for very high data transfer between it and important sub-systems such as graphics terminals, disk drives and high speed LANs.
The first to be defined was created by the Video Equipment Standards Association (VESA) and described as the VESA Local (VL) bus. It was an industry standard and the first products reached the market in 1993. However, its success now seems limited as there has been a significant swing to the second; the PCI bus. PCI is not an industry standard. That is, it is not open as are national standards. Rather, it was created by Intel who then passed the rights to a licensing body which includes Intel and others as shareholders. PCI, or in full, the Peripheral Component Interconnect bus is more general than VL and seems to offer more for future very high performance systems. Thus not only has it been adopted by virtually all the personal computer manufacturers, including Apple, but also by many of the workstation manufacturers. It will become a United States standard in due course.
The presentation was created using Borland C++ 3.1 and Asymetrix's Multimedia Toolbook 1.53. It runs under DOS on a standard 486 level PC (with Sound Blaster card). It requires a machine with VGA graphics.
In order to meet the objectives of the project, the data engineering specified that the text of the standards forms the core set of files for the presentation. They are modified slightly to include hyperlinks, but otherwise are as published. A similar file is included with these which basically describes computer busses in general and provides other tutorial information. The further files required for navigation are quite separate from this core.
Considerable attention was paid to the graphical user interface (GUI). A careful reading was made of the ergonomic principles used in related projects and incorporated here. When the user activates the interface the information is presented to them as a set of standards which are referred to as modules. Each module is divided into activities which describe a significant sub-section within the standard. For example, a description of the pin layout in the connector. Details of the presentation are too complex to present at this juncture, but will be demonstrated in the reading of the paper.
The first of these questions essentially defines the quality of the GUI and the navigational paradigm chosen. To answer this, a checklist was created for a target set of users and they were asked to respond. The checklist included questions along the following lines:
| Very satisfactory | 70% |
| Satisfactory | 30% |
| Neither satisfactory or unsatisfactory | 10% |
| Unsatisfactory | 0% |
| Very unsatisfactory | 0% |
The second is more difficult to test. A measurement was made by following a presentation with a multiple choice question test. This showed a mean score of 80%. Comparing the lecture time normally devoted to this material against the time students spent with this package would at least suggest the use of multimedia is far more productive and appears to achieve a similar learning outcome.
This paper has examined the particular case of computer bus standards. However, given the similar language of all standards, the approach adopted can easily be extended to others. One of its most important outcomes is that it is not especially difficult to adopt standards to multimedia and that new forms of presentation - such as animation - are well suited to the forms of the standards. As such, it allows the more widespread use of standard in engineering education and so familiarises students with a critical body of literature of their chosen profession.
Giacomo, J. D. (Ed) (1990). Digital Bus Handbook. New York, McGraw Hill.
Marcus, A. (1992). Graphic Design for Electronic Documents and User Interfaces. New York, ACM Press.
Qasem, I. and Mohamadian, H. (1992). Multimedia Technology in Engineering Education. IEEE Proceedings of the South Eastern Conference on Education, 46-49.
| Authors: Douglas Myers, Vei Ming Chong, Chun Che Fung School of Electrical and Computer Engineering Curtin University of Technology GPO Box U1987, Perth, Western Australia 6001 Please cite as: Myers, D. G., Chong, V. M. and Fung, C. C. (1996). A multimedia approach to disseminating engineering standards. In C. McBeath and R. Atkinson (Eds), The Learning Superhighway: New world? New worries? Proceedings of the Third International Interactive Multimedia Symposium, . Perth, Western Australia, 21-25 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1996/lp/myers2.html |