At first sight, it isn't clear that mechanical CAD software has changed much in the last five years or so. The CAD functionality of the leading brands is very similar. For basic mechanical design authoring, users making their next CAD purchase decision will find it difficult to differentiate between the functionality offered by the various vendors. Perhaps users can choose on price alone, or whether they liked the sales representative, or the marketing messages. Are we now into the mature phase of the CAD industry where brands determine purchases just like soap powder?

Market saturation is one indication of maturity. It is true that most of the people who need to have a CAD system have one already, even though you can still find localized areas of low saturation such as China or Eastern Europe. There are four well-established CAD software vendors each with a significant share who dominate the market. Most of the specialist niches, in evolutionary terms, have already been occupied-so, for example, we have Radan for sheet metal manufacture, ANSYS for simulation and analysis. The dramatic growth that characterizes the adoption of a new technology is a thing of the past.

Another characteristic of mature markets is that you would expect to find real standards-not the kind of de facto standards that exist because one vendor dominates the market. So, for example, you have GSM for cell phones or USB as an interface standard. Not so for computer software in general, nor CAD in particular. Systems functionality may be largely comparable but each software tool achieves this in unique, non-standard ways. Design data can be transferred between systems or translated into standard formats for data exchange or output processes, but this quite limited capability can be very complex to use.

In mature markets, the vendors find new customers by developing products with more features to sell to existing users of competitive systems. CAD products from each of the main players are capable of doing the job and differentiation in the market place hinges on relatively minor points of difference-another key indicator of CAD maturity. In fact, adding "bells and whistles" to CAD products that most enterprises don't want or need on a day-to-day basis does not improve design productivity. Users simply need a product that "thinks" in a way that suits their design process. However, all the added functionality has resulted in systems have become less easy to use.

Knowing the best way to achieve a particular result with a particular software tool is an important, but far less obvious, part of the investment that has been made. This is tacit knowledge that individual designers have but the organization probably hasn't captured. Can a designer who is competent on one system transfer to another and become proficient in a short amount of time? Maybe. Given the same task, would two designers, equally proficient in their use of the tool but working in different teams, achieve their results in the same way? Most unlikely-and this is a real problem. So there is a de facto "lock in" at work here. The typical user will upgrade rather than change.

So, is the CAD market mature? Yes and no! Basic CAD functionality is at a level where ever larger investments in developing existing systems will be necessary to deliver even relatively modest gains in productivity-the "s-curve" effect. It would be better to make a fundamental change in the technology, so that users can achieve a quantum leap forward in productivity terms. If the CAD story thus far has been a series of cycles, should we be expecting another one soon? At Cambashi we think the answer is yes, but it isn't PLM-or at least it isn't only PLM.

There are two quite distinct and very different ideas as to what PLM is all about:

Innovation-orientated PLM permits multiple versions of the truth and is about being creative and preserving options-until such time as the stakeholders agree upon the right trade-offs between options.

Control-of-change-oriented PLM allows for only one version of the truth.
Both can exist in the same enterprise. The key limitation of the control-centric approach is that if the project fixes the design too early in its life, creativity and innovation are stifled. However, both approaches involve a major shift in the design process and better integration of the design function into other business processes.

CAD has always been a very demanding application in terms of data storage. The data that define a single part are complex, consisting of many different but interrelated elements. Similarly complex is the way individual parts can be combined to form assemblies. When we try to hold a set of possible solutions, and the steps that got us there, and the simulations that evaluate them against our design criteria, we start to see even more data and more complex relationships. Much of the historic challenge for CAD has been about handling this complex data in unique and proprietary ways. While CAD has been following its own path, something very interesting has been quietly going on in the background-databases and middleware have steadily advanced the boundary that defines which tasks can be safely relegated to the "IT infrastructure."

We think that we are on the cusp of a new cycle in the CAD world-the internals are about to change. Ten years ago, mid-range modelers built on a higher platform of MS foundation classes, constraint managers, and kernel solid modelers emerged. They challenged existing CAD products that had to carry the overhead of coding their own user interfaces, graphics handling, and geometric modeling. Today, we are about to hit a new higher-level internal platform-one based on a standard SQL database handling features rather than a file-based, data structure-inspired heap. It will help handle geometry at the feature level rather than the part level and help maintain versioning and roll back. At the same time, building on top of standard middleware like .NET or Websphere will make interfacing with all those complementary software tools in the enterprise much simpler. These changes, in turn, will become enablers for the wider and faster adoption of PLM.

Within the enterprise, and more importantly the extended enterprise, the focus has shifted to realizing the creative capital of the whole team. CAD has been predominantly a tool for individual creativity for almost as long as it has been around. The complexity that permits different designers to achieve the same end result in quite different ways will have to be buried. Designers will need to work at a higher conceptual level in design terms, but within constraint systems that have been predefined. The next user interface paradigm is going to be one of designing in some kind of known context-for example, designing a mechanical assembly when the constraints of sensors for the electronic system are available to guide arbitrary design decisions.

So we reject the idea that CAD is a done deal. We await with interest the next cycle. We don't know if we should expect a skunk works product from one of the big four mechanical CAD vendors or if the next CAD product will come from some start up in Russia or India. However, we're sure it will come in the next year or so. You will just have to keep reading EAReport!

Bob Brown

bob.brown@cambashi.com

A version of this article was first published in the Engineering Automation Report, published monthly by Cyon Research Corporation.

Other Cambashi articles that may be of interest:

There's more to life than automated drafting tools for electrical design

Who will pay for the Building Information model?

Is PLM applicable to AEC?

back to top