Designing a new commercial airframe, military jet engine or launch vehicle is a task of enormous complexity. The creation of a multitude of unique new parts, the involvement of a legion of engineers, the need for frequent and intensive collaboration among them, the length of the development cycle, and the consequent expenses, are all on a scale that few other industries approach.

As an example, after several years of working quietly, Boeing formally announced plans for its new Sonic Cruiser airliner in 2001, with first deliveries forecast for 2008. Analysts estimate total development costs for this program at $9-12 billion. Delays in completing such a program can carry substantial financial risk, since early customer orders typically include penalties for late product delivery. Given this scenario, careful control of the cost and length of the development process is a top priority, and inefficient practices must be identified and eliminated.

One of these inefficiencies is the lack of comprehensive interoperability between the high-end mechanical CAD platforms used by aerospace designers. This problem has been a longstanding and unavoidable reality for these skilled professionals, stemming from the simple fact that the major high-end CAD solutions have evolved independently, with each having a proprietary data format. This has added untold millions of dollars and months of time to every large R&D program and has inflated product lifecycle support costs as well.

The CAD interoperability problem: why it is important

The need for CAD interoperability is a problem that cannot be mandated out of existence. Many large corporations have attempted to circumvent the issue by requiring that all of their design engineers use the same CAD package. However, this approach has never been particularly successful in the aerospace industry, where the community of design engineers working on a major R&D program is large and heterogeneous.

For instance, a prime contractor for a new military aircraft program may want to draw on the best in-house design talent, reaching across divisional boundaries when necessary. However, since most major government aerospace contractors have grown through mergers and acquisitions as the industry has consolidated, a wide diversity of legacy mechanical design platforms generally exists within each company.

Also, in today’s increasingly collaborative engineering environment, prime contractors routinely enter into design partnerships with key subcontractors in order to take advantage of the subcontractor’s design expertise as well as its manufacturing capabilities. These design partnerships form and re-form from program to program, with prime contractors and subcontractors often exchanging roles. In all probability, the “sub’s” engineering teams will utilize a different CAD platform than the “prime.”

Each time that a CAD file must be converted to a different format during the design of a part or assembly, the process can take days or weeks and cost thousands of dollars. Further, the design of a single mechanical part is likely to undergo numerous revisions from inception to the final production-ready version, with each enhancement to the design likely requiring additional CAD data conversions. This file transformation activity adds little inherent value to the quality of the final product, while stretching out the development process and adding significantly to its cost.

An added complication is the need to service, repair, and overhaul sophisticated aerospace products during a life span that can sometimes be measured in decades. For example, formal development of NASA’s space shuttle started in 1972 and took nine years to complete, a design that continues to be used today, 21 years after inaugural flight. Accessing old design data, years after a part was created, can be a major challenge. Sometimes, the design software used to create the original part is obsolete. Often, only an original Mylar drawing or, at best, a geometry-only computer file based on an industry standard are the only design data available. In either situation, an engineer must manually recreate the design in a modern CAD system so that a part can be modified or an electronic file sent out for bid to suppliers. A means of storing complete CAD model data in a durable format that is not CAD-vendor specific is the most viable long-term solution to this problem.

The creation and maintenance of standard parts libraries is another area where poor CAD interoperability causes inefficiency in this industry. Although these standard parts tend to be relatively simple, they are custom-designed, and a single type of part may be used hundreds or thousands of times in a completed airframe. A large aerospace company that has grown through mergers and acquisitions is likely to have numerous libraries of standard parts — created and maintained using different CAD systems — scattered around different engineering groups or divisions. This lack of integration leads to many cases of design redundancy, where the same part exists in different CAD formats in different libraries.

When such redundancy exists, many people must work to keep each of these libraries up-to-date; and, more significantly, duplicate inventories of parts must be kept on hand. While the problem may sound trivial, it is important to consider that a single commercial airliner can easily contain more than a million of these simple fasteners and brackets. A way to maintain a single, centralized parts library in a “universal” data format could eliminate this waste and save millions of dollars per year.

Existing solutions and their limitations

Existing approaches to achieving design tool interoperability have run the gamut from fully manual methods to solutions developed in-house, along with the ongoing, parallel evolution of industry standards. While each has provided some benefit, none has effectively eliminated this major bottleneck in the engineering process.

The problem cannot be solved simply by the “brute force” technique of throwing more manpower at it. Hiring legions of consultants to be available to rebuild models on demand is cost prohibitive in today’s economy. Aerospace companies are looking for smarter ways to develop products, while still reducing manufacturing costs.

Numerous commercially available solutions exist for translating CAD files; but, until recently, these were only capable of dealing with basic design geometry, a derivative subset of the full information that specifies a design. A variety of products, tools, and outsourced services are available to interchange this geometric information between the major CAD applications now in use. However, their common limitation is their inability to exchange parametric features and related design history and constraints.

These higher-order features, history, and constraints — collectively termed “design intelligence” — are the vocabulary of the modern design engineer. When design intelligence is lost in the process of exchanging CAD files, the receiving designer loses a complete insight into the intentions of the original designer. Often, it becomes necessary for an engineer to re-create this design intelligence manually, through a process called “remastering.”

Industry standards also have played a role in addressing the CAD interoperability issue. IGES (Initial Graphics Exchange Specification) and STEP (Standard for the Exchange of Product Model Data) have both seen some acceptance in the aerospace industry. However, while useful for some CAD data exchange purposes, these formats still suffer from the “geometry-only” limitation, with the same requirement for manual remastering of translated files. In fact, the STEP standard committee has recognized the importance of feature-based exchange–and is moving in that direction. However, an enhanced standard is still years away from ratification and even further away from commercial implementation.

Occasionally, companies will attempt to address the problem by developing their own in-house CAD translation solutions. Most of these efforts fall by the wayside, as their sponsors discover the magnitude of the task and the constant need to update their tools to reflect the ongoing improvements to commercial CAD software packages. At best, they find use as one-time translation facilitators for companies that have elected to migrate from one CAD package to a different one.

In addition, none of these solutions provides users with the capability to store feature-level CAD data in a CAD-package-neutral format for the long term. Lacking this, some of the more important CAD interoperability requirements of the aerospace industry cannot be addressed.

Feature-based solutions to the rescue

Within the last two years, a new breed of CAD interoperability solutions has reached the market. For the first time, it has become possible to exchange full design intelligence between disparate CAD platforms, cutting time to market and reducing development costs significantly. The most advanced systems provide other business benefits that go beyond the efficient interchange of CAD files. One example is the Collaboration Gateway from Proficiency, Inc., Marlborough, MA.

The Collaboration Gateway interoperability solution is a stand-alone software product for large aerospace OEMs and suppliers. This web-based application operates on a company’s internal IT infrastructure and consists of a browser-based client, an application server, and “design agents” that co-reside on existing CAD workstations. Because it is typically installed entirely within a company’s firewall, job turnaround time is quick, and confidential design information remains securely in-house.

The application operates automatically and without user intervention. This capability permits the user to “batch up” numerous jobs to run overnight or on weekends, taking advantage of periods of low computer usage while freeing up the designer’s day for more creative work. The user can be notified by e-mail when each task is complete.

Proficiency’s software system also creates a detailed set of reports for each task that is submitted to it. These reports allow the user not only to see task results at a summary level, but also to “drill down” and understand how an exchange was carried out on a feature-by-feature basis. Further, the application calculates and compares the physical properties of a part in both the “source” and “target” CAD systems to provide additional assurance that the design has not been altered inappropriately during the exchange process.

At the core is the company’s Universal Product Representation (UPR) technology. Because the UPR is a persistent store of CAD model data in a neutral, independent format, it is uniquely able to address the project’s CAD interoperability needs in the post-introduction phase of an aerospace product’s lifecycle, including the requirements for support engineering and part library consolidation and maintenance. The software currently supports the high-end CAD applications most commonly used in the aerospace industry: Pro/ENGINEER 2000i2 and 2001, I-DEAS 8 and 9, Unigraphics 16 and 18, and CATIA V4.

Users have seen file exchanges accomplished in as little as one-tenth of the time required for a geometry-based solution with the requisite manual remastering, so the benefit it provides is clearly compelling. Though it is too early to calculate definitive ROIs, these same users estimate that this software can shave months of time — and millions of dollars — from a typical aerospace development cycle.