Design News for Mechanical and Design Engineers

Maybe you should worry. Engineering wages in India, for example, pale in comparison to ours. "In India, we can hire a beginning mechanical engineer for $12,000 and a Ph.D. for $20,000 a year," says Tom Epply, president of Continental Design, an outsourcing firm with access to about 200 Indian engineers. Compare those wages with those in the U.S., where the average wage for a mechanical engineer came to $67,430 in 2003, according to the U.S. Bureau of Labor Statistics. Even after you factor in costs related to extra training and managing a distant engineering team, that engineer in India will still cost a whole lot less than his American counterpart. "In general, for every engineering resource here, we can get three in India for the same amount of money," says Seshu Seshasai, vice president of technology for Textron Fastening Systems and one of the executives charged with creating an Indian technical center for all of Textron's business units.

When confronted with such compelling cost arguments, some companies will blindly chase the cheap labor. Yet it may pay for them to be less reactionary about outsourcing. Think of it as an elective surgery. If done the wrong way, for the wrong reasons, or at the wrong time, it could turn out very badly and cause lots of pain. But if done right it may just work out for the best. If you work as a design engineer, "best" boils down to two things—making sure the work you do isn't the kind that can be outsourced easily and making sure the products you design don't end up in the wrong manufacturing environment. Here's a look at both issues:

One thing that increasingly looks susceptible to outsourcing is, well, your job. India, in particular, is currently a magnet for companies wanting to draw on low-cost mechanical and electrical engineers. Outsourcing firms already employ thousands of Indian engineers. One such firm, QuEst Inc., has 600 engineers in India, plus 150 more in the U.S., Europe, and China. "The technical talent in India is top-notch and the rates are a huge advantage," says Kurt Noe, QuEst's business development director and a veteran engineer who previously directed Hamilton Sundstrand's research programs at the United Technologies Research Center. Large technology companies—including GE, Honeywell, and Textron—have over the past few years opened their own engineering and research facilities in India, too.

Some of the work done in India does overlap with what the most advanced engineering teams do here. At a press conference in New York last year, GE Advanced Materials Technology VP William Banholzer reported that GE engineers and scientists in Asia engage in advanced research and development activities. And Continental's Epply, who had a 30-year engineering career at General Motors, claims that there's almost no product development work he wouldn't send to India with confidence—with the possible exception of tooling and fixture design. "The only other thing we can't send overseas are certain government contracts," he says.

For the most part, though, companies interested in outsourcing take a more measured approach and keep their core new product development work in house. Epply, whose customers include Delphi and other automotive suppliers, notes that Indian engineers don't typically get to work on new product development, at least not right away. "They have to earn their way by working on all the mundane stuff first," he says.

And that's a good thing, in Noe's view. As someone who once purchased engineering services for United Technologies before joining QuEst, he advocates finding a "globally optimal mix" of engineering resources. The core engineering work, the kind that tends to involve proprietary knowledge or intellectual property, would remain in house while outsourcing firms take on the non-core work. This latter category includes some simple tasks, such as converting drawings or working on legacy products. Increasingly, though, it can also mean more advanced work, such as running finite element analyses, conducting product tests, or handling some project management. The "smart way to do things," according to Noe, is to free up good design engineers to do what they do best—create new products. "If you have an engineer in the U.S. with 20 years experience and deep domain and product knowledge, he is of greater value working on new product development than sustaining legacy products," he argues.

Textron recently followed this global optimization strategy when it opened its technical center in Bangalore, India late last year. A shared resource for all the Textron business units, the center has space for about 300 engineers and will have roughly 225 working there by the end of this year, according to Seshasai, who served as chairman of a technology council that includes the engineering leaders from all of the company's business units. "We decided we needed a presence in India to exploit the engineering resources there," he says.

Large technology companies are scrambling to take advantage of India's burgeoning engineering market. Textron has opened a new technical center with space for 300 engineers, two of whom are shown here.

The company's process for deciding which tasks to perform in India went something like this: Seshasai and other technology managers first identified all the technical skill sets and day-to-day engineering tasks involved in their businesses. They then entered this information in a matrix that helped them get a sense of which competencies and tasks were both important from a technology standpoint and difficult to perform elsewhere (see diagram at left). This core work remains in house. Work that was neither core or hard to do elsewhere has or will soon go to India for substantial cost savings. And work that fell somewhere in the middle is now a candidate for automation and other process improvements to make it more productive.

Textron Fastening Systems' "non-core" work has so far involved tasks such as converting 250,000 paper and CAD fastener drawings into a common format and creating a database of all these fasteners based on 22 dimensional parameters. QuEst's Noe notes that Indian engineers have a threefold advantage in these kind of tasks. For one thing, they cost less than their American counterparts. For another, they may focus almost exclusively on this kind of repetitive work. "They end up being more productive than an American engineer because they do it over and over again." And finally, this kind of work is not the type that U.S. engineers here really want to do as their careers progress.

Textron's roadmap for its Indian engineering center doesn't stop at drawing conversion. Seshasai reveals plans to use the engineering resources there more fully over time. He estimates it will take another six months to fully integrate the Indian engineers into Textron's engineering teams and corporate culture. After that, he plans to let them work extensions to legacy products. By the third or fourth year, he expects that the Indian engineers will "be doing the same kind of work that we do here," including new product development.

Even then, though, a significant number of Textron Fastening's engineering jobs still won't lend themselves to outsourcing. As Seshasai explains, his business and many others that offer custom-engineered products, require hands-on application engineering. "Half of my engineering team has to interface with customers," he says. Those engineers, who also generate many of the new product ideas and have the most in-depth knowledge about fastening, will have to stay near their customer base here and in Western Europe. In addition to application engineers, Seshasai makes a case that other "cross-functional" engineering roles also tend to resist offshoring. He puts project managers, product managers, and sales engineers in this group.

Seshasai and Noe's views hint at some good career advice for any engineer worried about offshoring. Noe says that deep product knowledge and ability to work with or create intellectual property will leave you in better standing than more mundane work. And Seshasai stresses the importance of developing the project management, communication, and customer relation skills that are more difficult to cultivate in the low-wage countries. "Engineers today need to be multidisciplined to get the most out of their career," he advises.

Manufacturing: Should It Stay or Should It Go? Politics aside, the real issue surrounding offshore manufacturing is not whether products can be made more cost-effectively overseas. Some clearly can. For example, so much of the infrastructure for making displays, notebook computers, and many other types of electronics has moved to Asia that offshore manufacturing sometimes represents the only practical choice. The same goes for many toys, hand tools, and labor-intensive consumer goods such as apparel. A quick walk through the aisles of Wal-Mart would turn up scores of other examples. And even in the absence of lower costs, companies make strategic decisions to set up shop overseas—for the sake of proximity to their corporate customers or growing overseas consumer markets. On the other hand, some products resist offshoring. Examples are products produced with so much automation that they require little labor. Some military products or those with export controls likewise resist offshoring. Many products, though, still fall into a gray area in which plausible cases can be made for both domestic and offshore manufacturing. Which ones shouldn't go overseas? Nick Dewhurst and David Meeker believe they have some answers. They recently published a detailed report on the cost of offshore manufacturing in China, though their conclusions apply to any low-cost labor market. The report draws on Dewhurst's experience as the vice president of Boothroyd Dewhurst, where he works with OEM clients to streamline their product designs. Meeker, meanwhile, has spent more than 20 years in design engineering, working for companies like Digital Equipment Corp. and Compaq. Nowadays, he consults on product design and development and teaches at the Massachusetts Institute of Technology. Both men believe that engineers and purchasing staffers often make decisions about offshore manufacturing without having all the facts they need to make the best decision. "We're not trying to stop offshore manufacturing, nor could we," Meeker says. "We just believe that engineers don't always understand all the costs involved with sending a product overseas." Many companies chase cheap labor without recognizing that labor can make up just a small fraction of total manufacturing costs for many electro-mechanical products, according to data from Boothroyd Dewhurst. As an important example of this misunderstanding, they point to a widely held fallacy about the low labor costs that lure many companies overseas in the first place. "Many companies attach too much weight to labor costs," Dewhurst says. But these costs may not represent as much of the total cost as you might think." Looking back on the jobs Boothroyd-Dewhurst has worked over the past decade, Dewhurst estimates that the labor cost for a typical electromechanical assembly would be just four percent, versus 72 percent for part and material costs and 24 percent for overhead. He believes similar numbers hold true for many kinds of consumer goods, particularly those made on automated lines. Sure, labor costs can represent a far bigger portion of total for other types of products. "But they rarely represent the biggest chunk of the cost," he continues. He and Meeker go on to detail a number of often overlooked costs associated with offshore manufacturing, including the cost of shipping and shipping losses, travel, communication, and vendor selection. Then there's the cost of lost flexibility in dealing with engineering or marketing changes. "By the time a change comes through, you may have your entire inventory sitting on a ship in the middle of the ocean," Meeker says. To these costs, they add higher prices of any materials not readily available in overseas markets. And they go on to list a host of costs that don't directly go into the calculations of offshore manufacturing costs—but should. These include costs associated with theft, piracy, legal work, training a new workforce, and any layoffs in the U.S. Offshore manufacturing also produces some slippery costs related to cultural issues and the poor morale outsourcing can engender for the shrinking workforce back at home. Finally, there's the cost of poor quality. Meeker acknowledges that plenty of Asian manufacturing operations can offer quality comparable to U.S. and Western European levels. But some simply cannot or require very close supervision to do so. Ensuring high quality thus becomes a matter of "constant vigilance," Meeker says. "Companies happy with offshore manufacturing have poured a lot of time and money into quality," he says. "The successful ones have been at it a long time." Some quality issues rear their heads in ways that are easy to quantify. For instance, reworking the ejection system on a poorly constructed injection mold takes a fixed amount of time and money. But Dewhurst also points to more subtle effects from poor manufacturing quantity. "Products that don't work the way they are supposed to can cost OEMs their good reputation," he says. When Dewhurst and Meeker tallied up all the easily quantified offshoring costs, they discovered that the offshoring adds increase total manufacturing costs by at least 24 percent. "And that's a conservative number," says Meeker says, who notes that this figure includes only tangible costs related to logistics. In most cases, the figure would be even higher. For instance, the report contains a case history in which one of Boothroyd-Dewhurst's customers, a consumer goods manufacturer whose name the authors wouldn't reveal, had a cost add-on of close to 30 percent for their operations in China. The lion's share of offshoring costs may be beyond your control, but design engineers can still influence the offshoring decision in a couple of ways. Meeker advises design engineers to pore over the bill of materials for any products that may be destined for offshore manufacturing. "To make a good decision about where a product should be manufactured, it's important to make sure the BOM has the right information," he says. One error he's noticed in his consulting work is that some corporate computer systems will fill in an estimated price for a component or material based on past experience here in the U.S. "That same component or material could cost far more overseas," he says. Another way you, as a design engineer, can influence the decision about manufacturing location is to design economical products. Dewhurst and Meeker both say that some of the products now being made overseas go there because they have inefficient designs from a manufacturing or assembly standpoint. The consumer goods case history mentioned in their report involved an existing 26-component elec-tromechanical subassembly manufactured in China. A design-for-manufacturing-and-assembly analysis (DFMA), using software and methods developed by Boothroyd-Dewhurst, reduced that subassembly to 12 parts. "It's rare that you can't take a big chunk of labor and material out of a product once you apply the DFMA methodology," Meeker says. In this case, the cost reduction came to 27 percent. And if you add that to the 24 percent offshoring cost, "the product would actually cost less to manufacture in the U.S.," Meeker adds. One flaw that seems to mar Dewhurst and Meeker's results involves the portability of design work. There's nothing to stop companies from taking a design optimized with DFMA methods and shipping it over to a low cost labor market for production. "That looks like the gaping hole in our argument," Dewhurst acknowledges. But is it really? Even an optimized design would still be subject to the extra offshoring costs that Dewhurst and Meeker lay out in their report. Just to break even when manufacturing overseas, the product has to overcome that 24 percent add-on, Meeker explains. So the lesson to take away is not that it never makes economic sense to manufacture in low cost labor markets. Sometimes it might. But you won't know without some extensive research into costs and an effort to create economical designs. And Dewhurst lays much of the responsibility here at the feet of design engineers. "Don't automatically buy into the underlying assumption of outsourcing," he says. "Low cost labor doesn't always equate with low cost products."   What Engineers Need to Know about Outsourcing Don't panic. Some engineering jobs have gone overseas, but many engineering functions don't lend themselves to outsourcing. Focus on the core. Engineering activities related to your company's new product development and intellectual property are the most resistant to outsourcing. Soft skills matter. Communication, project managment, and customer-handling skills are harder to duplicate overseas.