The AutoCAD pump assembly design is called an "assembly model" and the component designs are called "component models." These AutoCAD models are actually files on Peter's file server. The assembly model contains pointers to the component files. The vendors supply AutoCAD component files for the off-the-shelf parts. Thus, the designers don't need to re-draw the off-the-shelf components, and components that have multiple instances in an assembly or in multiple assemblies are stored just once. Naturally, updating a component model file updates all assembly models that utilize it.
When the assembly model is printed on paper (usually several sheets), it is called a "drawing." Likewise with component models. Some day the manufacturing shop will view models directly on computer screens, but for now assume the shop works only from paper.
When the shop or customers find design flaws in a pump, the design model is updated and assigned a revision number. The first issue of the design has no revision identifier (null) and the first revision is identified "A", the second is "B", and so on. The shop needs to know that the paper it is working to is the latest revision.
Peter has hired an embedded systems programmer to write the software for the pump control system. The software goes through version revision on a regular basis. Peter wants to control both source and executable files on his file server.
This is not an easy exercise, but it is not particularly difficult either. This exercise distills the essence of some actual data modeling problems. You might find it easiest to draw a general high-level model chart first, then think about the problem for a day or two, and return to the model chart and revise it. This might also be a good exercise for a small team working over several days.
Email: Richard dot J dot Wagner at gmail dot com
dm09.htm, this hand crafted HTML file created October 18, 1997.
Last updated October 22, 2010, by
Rick Wagner. Copyright © 2010, all rights reserved.