What is your Philosophy of CAD Standards?

[cadtag]

One of the questions that seems to get swept under the rug whenever an organization starts developing their standards is the "why?" of what they are doing. There's an implicit acknowledgement that standards are a Good Thing (TM), and that an organization ought to have some, but there rarely seems to be much discussion about what the philosphy behind the CAD Standards is, and what the standards are going to accomplish.

This is unfortunate, because the shape and implementation of a CAD standards effort is wholly dependent on what the ultimate goal is. Standards and processes that are devloped to meet one need are often in conflict with other desirable goals. A standard that is focused on immediate production of similar drawings within a small office that deals with a limited number of clients/agencies is probably not going to be that useful in a multi-discipline, multi-company, multi-continent environment. Similarly, a standard that's focused on AEC and multi-story office buildings is not really going to be very user-friendly to a CAD group focused on ALTA surveys or roadway design. And even less useful to a firm that does

So, what's the philosophy behind your CAD standards? What are you wanting to accomplish? What aren't you concerned about? What works for you -- what doesn't? And, what kind of product do you use CAD to design?

[duhvinci]

My purpose for establishments and on going modification (really important) of standards is to facilitate in the efficient production and subsequent automation of final drawing production. Allow me to elaborate.

Firstly my industry is custom machinery.

Secondly I perform contract work for multiple customers.

To address the second consideration first. I established my own set of standards to allow me to work with only one set rather than switching between multiple sets requiring adjustment inefficiencies. By strict adherence to my standards drawings can be automatically modified to the customers standard when complete.

When I created my standards I tried to ignore convention at the time - which by the way are still pretty much dominate.

The established convention at the time was based on the following factors which no longer are considerations: Limited layer counts, limited monitor colors, limited computer resources, etc.

For me

Layer names of Hid or Cen or Phan and derivitives were out.
Linetype and color by layer was out. Each entity was to be set individually.
Each of those were nonsensical to me.

The machines I design are made up of multiple components, it made no sense to have layers that did not reflect individual components so each component has one layer. Those objects that make up that component (object, hidden, center lines, all exist on the related component layer) In this way going from 3d model to 2d detail could easily be automated.

Oh, and you make some very good points about the usefulness of a small operation's (like me) standard for a big conglomerate. But that is the fault of the ingrained bureaucracy within the larger structure. Believe me I tried exhaustively to influence my clients methods years ago and it was like running into a brick wall. I experience the same frustration in attempting to promote the benefits of my software. This inherent refusal to make change is one of the reasons the US has lost so much of its competetive edge. <off soapbox)