Interoperability is Still the 3D Elephant in the Room

Ken Tashiro, Vice President and COO, Elysium Inc.The famous Tower of Babel story tells of a brazen human attempt to build a city with a tower that had its top in the heavens. Not willing to tolerate humans trespassing in this celestial domain, God decides to punish such impudence by confounding the speech of mortals, thereby preventing them from working together to complete the great project. So ended Babel's ambition, and so began the confusion of human speech with which we live today.

In the world of aerospace design and manufacturing, we are in a similar predicament. Amazing tools have evolved that continue to enhance our ability and our ambition to reach the heavens – or at least build products that can. Yet, by a process that we seem reluctant to recognize or acknowledge, our crucial ability to share data and interoperate continues to be compromised by proprietary coding, algorithms, and other built-in assumptions among competing toolsets. In short, like the ancients of Babel, the speech by which our design and manufacturing systems communicate is forever being confounded with business implications that may grow only more serious with the passage of time.

Modern design, development, and production are deeply intertwined with the tools of CAD and, as model-based definition grows in importance, there is an implicit assumption that critical design data will move accurately and smoothly into each successive application, as ideas transform into product. However, the current tools of the trade, prepared by different vendors and issued in different versions, are often built around subtly different assumptions, algorithms, and formulas, and then coded in proprietary formats. Worse still, this fact is not always obvious.

Often everything seems like it will work together. However, the introduction of different design and production tools to the same set of input data might render end products that are substantially different in form, fit, and finish. Or, more often, somewhere along the way these glitches accumulate to the point where something in the process simply does not work, and expensive manual intervention by design engineers, technicians, and programmers is needed to make things right.

For the aerospace industry (even more so than for automotive, consumer goods, etc.) these problems are becoming intolerable. Most fundamentally, quality and functionality are more critical in aerospace than in almost any other industry. With human lives at stake, the consequences of failure are completely unacceptable. In addition, unlike automotive, for example, product life cycles are longer, often measured in many decades. The implications are that the data for making a part in 1999 must still be valid and actionable by current systems in 2019 and 2049, or beyond, no matter how many upgrades of a given software product have been deployed and regardless of whether the software platform has been changed.

This is one of the most critical areas for those involved with design engineering today. If enterprises are to communicate with their suppliers, if the value chain is to function to its potential, and if products are to stay safe, reliable, and cost efficient, it is of paramount importance that interoperability should be built-in, something we can count on: an a priori assumption. The rich, parametric intelligence inherent in CAD models must be seamlessly transferable from one CAD platform to another.

Where We Are
In reality, of course, we are a long way from that goal. The fact is that the entrepreneurial and highly competitive design automation market has spawned dramatic advances in capabilities as rapidly as it has created invisible barriers – proprietary coding, calculation, and transmission methods – that militate against assured data movement from platform to platform. In fact, one almost guarantees the other. The companies that invest so much in advancing the state of the art quite naturally do not want to fully open up or give away their core technology. Moreover, they are striving to create integrated, native suites, which extend from CAD and CAE to CAM and beyond, and see a real benefit in having customers commit to their individual platform or technology. Yet, industry will always choose what it considers best of class and try to forge the other segments into a family of workable applications.

Aerospace firms have grown adept at finding workarounds for translation problems. In addition, there are promising efforts to develop universal standards for sharing and archiving data. Nevertheless, these efforts are tentative and may not be adequate to meet the vastly expanded challenges of an emerging world in which model-based definition – leveraging 3D CAD – becomes the standard by which suppliers create parts and assemblies.

Geometric dimensioning and tolerancing (GD&T) will need to be embedded, when necessary, and reliably shareable so that data can propagate throughout the full life cycle of design changes.

Here again, at present, there are few certain ways to translate all this critical part data from, say, a Unigraphics file format to something usable with CATIA. Too often, these translation gaps are filled by the manual rework of CAD files, the cost of which is absorbed in various ways by the whole industry. When the model re-creation is incomplete, or includes errors or misses some key bit of definitional data, it means even more work and more cost.

Even version control can be a nightmare due to these underlying data issues; backward and forward compatibility within a given environment is never something to take for granted. Every time an organization implements a new version of software, they need to travel a learning curve to verify compatibility with other systems.

No industry can afford this. Aerospace is perhaps the most exposed but automotive and consumer and industrial product companies face the same issues.

Looking Ahead
Fortunately, some efforts are being made to address the interoperability challenge, notably the Boeing/Northrop Grumman Global Product Data Interoperability Summit. The second such event, held last November in Arizona attracted a substantial cross section of the industry (with the third scheduled for Nov 7 - 10, 2011). Wisely, event organizers have insisted it be open and agnostic – an opportunity to share common ideas and solutions about data interoperability challenges.

In a statement, Doug Norton, vice president and chief information officer, electronic systems sector, Northrop Grumman Corporation, underscored the seriousness of the work ahead, noting, "Product design and development quality and success, in any industrial market segment, are increasingly dependent upon mastering data exchange management across the enterprise."

The Boeing-Northrop Summit is a good start. Aerospace organizations need to participate in these initiatives while making efforts to identify the scope and impact of this problem. This is for their benefit, and so they can share the ongoing importance of the issue with their vendor community. Of course, technical efforts must continue in an effort to bridge gaps and strengthen interoperability whenever possible, including further fine-tuning of solutions that reliably translate data between different tools.

If we are to avoid the fate of Babel we must give this matter the attention it deserves. What is needed is knowledge, awareness, and a collective commitment to change.

During World War I, the French premier, George Clemenceau, observed that, "War is too important to be left to the generals." In our rapidly evolving world, data interoperability is too important to be left to the vendor community. It is everyone's problem and we all need to be part of the solution.

Elysium Inc.
Southfield, MI
elysiuminc.com