Normally, product or component development and engineering involves, at the very least, designers, engineers, and detailers. Each of these individuals or groups works in a sequential fashion, often with feedback loops meaning that sequences once done will be sequences modified and repeated.

Stephen Kennel, president, Virtual Engineering, Inc. (Plymouth, MI), recalls that when he was doing engineering at Johnson Controls Automotive operations headquarters (Plymouth, MI-small world), it was typical that he would provide input to a designer who would, in turn, come back to Kennel with his take on Kennel's instructions. It was rough going. Often, it was necessary to go back to the proverbial/actual drawing board as there was a lack of congruence between the designer's and the engineer's view of what should be what.

Kennel believes that the right way to do this is to give the project to an engineer, who is fully equipped with a workstation and software (in the case of Virtual Engineering, the tools of choice are Silicon Graphics workstations and Pro/ENGINEER software from Parametric Technology) and the charge to get the job done. And then the project moves forward with a single point of responsibility and a minimized possibility of redundancy.

"Usually," Kennel says, "the CAD designer in a conventional engineering firm or manufacturing company is a non-degreed individual working on wireframes. We have degreed engineers working on solids. We believe in the concept of 'engineer as designer.' Which means that there are a minimum of hand-offs."

He explains that the engineer at Virtual Engineering is responsible for doing everything. This means the design, the analysis and even the detailing. "Accountants argue against the use of the engineer as a detailer," he admits. But then he goes on to explain that when considered from the point of view of the entire project, this comprehensive involvement of the engineer is beneficial because there are fewer iterations involved because the possibilities of misunderstanding are minimized. "We end up with a more robust design."

And as for the cost situation, he points out that 95% of the projects that the company has done are on a fixed-cost basis, which, he says, "puts the incentive in the right place": the Virtual Engineering engineer has no reason to drag his heels or to milk the job. (What about the other 5%? 'Well, it would be 100%," Kennel answers, "but the accounting systems at these customers can't handle the fixed-cost approach.")

According to Kennel, he helped establish Virtual Engineering in 1992, having left Johnson Controls Automotive, where he had been director of Engineering. He says that as they surveyed the scene back then, there were essentially two types of contract engineering firms: the large, veteran Detroit firms, which he says have plenty of non-degreed people working on wireframes, and specialized engineering companies performing analysis and simulation. So they decided to start a different type of engineering firm, one that would have the latest in design and engineering technology and the people who can make it buzz. (Kennel is quite bullish, for example, about solids-based design, which he describes as "an underestimated tool." He explains, "It provides a total description of the part. You can run tolerance and interference checks. You can understand the center of gravity. . ." Briefly: he thinks it is the design/engineering tool of choice.)

Although Virtual Engineering has done work for a variety of Tier-One suppliers and OEMs-Lear Corp., ASC, Webasto, ITT, Bosch, Detroit Diesel, Johnson Controls, TRW, among them-Kennel says that the company's approach to engineering services is a real boon to the Tier-Two and Tier-Three suppliers, who may not have the resources to maintain a full-time engineering group or simply to attract the kind of talent necessary for the projects that have become more demanding now that the OEMs are pushing more of the design and engineering work down the food chain. "They can use us when they need us. We're not adding to their overhead." He adds, "And they can clearly assign the cost of engineering to the particular project."

Is there a time savings that accrues from this approach? Conceivably, but not necessarily. "The length of a program is a function of how it fits into a customer's program," Kennel says. That is, although Virtual Engineering could conceivably do their portion of, say, a seating system in a short period of time, it depends on their customer's program length (which is a function of their customer's customer's requirements), as it would not be advantageous to have their job done before it was required.