Last year in March, the National Institute of Standards & Technology (NIST) released a landmark report that gave us the first in-depth look at the cost of achieving computer-aided design interoperability. Called the “Interoperability Cost Analysis of the US Automotive Supply Chain,” the study estimated that “imperfect” data transfer between CAD systems imposed a penalty of “at least $1 billion per year” on the automotive segment of the nation’s design and manufacturing industry.
The report concluded that of the two main methods for improving interoperability-developing software for direct translation of data from one system to another or implementing a standard neutral format for data exchange-the latter approach, specifically adopting ISO's STEP (STandard for the Exchange of Product model data), was the most feasible. As evidence, the study cited numerous success stories involving STEP from leading automakers and suppliers. It also announced that NIST would develop tools to further STEP's acceptance and use.
In the 14 months since the report was published, many new ideas have been advanced as potential solutions to the interoperability problem, including new Web-based collaboration tools and services that promise direct data exchange between users of diverse CAD systems. As a result, some vendors and users appear to be curtailing their STEP efforts in anticipation of these new technologies. In fact, at the recent Daratech CAD/CAM/CAE Strategy Workshops-one of the best places to spot trends in the industry-STEP was discussed much less frequently by both user and vendor presenters than at any time since the first STEP specifications were standardized by ISO in 1994.
One exception to the rule at the Daratech conference was a presentation by Suresh Pillay, mechanical engineering automation manager at TRW's Space and Technology division, who argued that the industry should move forward aggressively with STEP. Pillay's situation is representative of many users who seek an edge in highly competitive markets. "We tend to push software envelopes to gain every advantage in reducing our cycle times," he notes. Thus, even though TRW has built its MCAD process around a single system-Dassault's Catia-Pillay says his division also uses "just about every CAD system on the market."
Consequently, TRW engages in massive amounts of data exchange. Though Pillay's team members do some manual translation, in about 95% of the cases the process is automatic, thanks largely to TRW's insistence that its partners who don't run Catia must use STEP. "This policy took enormous effort to put in place," Pillay says. "And every day people say the standard doesn't work. But it does. You just have to put some effort into it."
Some would argue that a technological bullet will come along, perhaps from a Web-based application service provider, that will enable direct translation and simply blow the interoperability problem away. Others contend that CAD vendors will eventually open up their geometry to allow direct translation between systems, especially now that their proprietary definitions-the very cause of interoperability snags-are no longer a market differentiator. But there is no evidence that either scenario is imminent or even likely.
Pillay, for one, maintains that the only way to improve interoperability is to keep insisting on standards. In fact, this appears not only to be the most reasonable course for the present, but one that will be even more critical over the next 12 to 18 months. As new CAD and collaboration tools enter the market at an ever faster pace, the importance of seamless interoperability will rise accordingly. Now is the time to pursue STEP more vigorously, not adopt a wait-and-see attitude.
Phil LoPiccolo: Editor-in-Chief