During the panel discussions on ‘Proprietary lightweight 3D formats and standards: A collision course?” at the 2008 CIC Conf
that Longview Advisors hosted earlier this month in Denver many of the audience’s questions and the resulting discussion touched on ‘what really is a lightweight 3D format for and when should you use it?’ In thinking about it and talking with some of the other folks on the panel, it occured to me that it would be useful to look back at why the lightweight formats were introduced in the first place and how the external constraints driving the requirements for ‘lightweight’ have changed over time.
Historically, the primary use case driving the creation of lightweight formats has really been about collaboration, and probably the most simplistic type of collaboration at that: viewing of product data by a third party. Prior to moving to a 3D process, this was achieved by creating fully annotated, physical 2D drawings and shipping them whereever they needed to go. Today, several of the world’s leading manufacturing companies are struggling right now as they attempt to shift their processes away from 2D drawings in favor of sending fully annotated digital 3D models throughout their supply chain and as more and more of the manufacturing industry follows their lead, they too will hit the same roadblocks. Sending source CAD data doesn’t work because
- CAD files are too intelligent as anyone with the source CAD assembly data can do whatever they want with the designer’s IP
- Asking everyone in your supply chain to buy a CAD seat to view data doesn’t sound so good when they used to receive drawings for free
- CAD files are too big (which is another way to say WAN network throughput is too slow and expensive)
And so many different ISVs independently tackled the problem of how to provide for free viewing of CAD data created with their tools and today you can choose from JT, 3DXML, DWF, eDrawings, XVL, and many more lightweight proprieary formats and their associated viewers. At first most of these formats were simple tessellated data organized in a node structure that mimicked the CAD assembly product structure. The key phrase here is ‘at first’ which brings us to the question of what changed along the way, which I will save for another time.