It's a new world out there, much of it moving across the Internet. The 681 respondents to the 1998 survey were asked, "Are you using the Internet more frequently for collaborative design?" Almost half of the American and Asian engineers replied, "Yes."
The engineers from both sides of the Pacific were asked to elaborate on how the Internet has changed the way they work. Their replies are instructive.
"It allows team members in different physical locations and time zones to more efficiently coordinate efforts," a Silicon Valley systems engineer notes.
"I can do business anytime, anywhere," an Asian engineer writes.
"The Internet makes it easier to obtain vendor and component information that is current," a California design engineer observes.
"There is always an expert to answer a question," a New York software engineer adds. "It's a huge technical resource."
It is good for dealing with people outside the company. "Extranetting remote contributors, greatly enhanced the effectiveness of their contributions," replies a Californian.
Collaborative design over the Internet, says a program manager, "removed traditional barriers for getting/sharing information. It helped build on others' ideas and solutions."
Internet: Collaboration on design
(Figure)
Some use it for inside work, like a Texas design engineer who says, "It has provided a convenient place to post user documentation and internally developed tools." But he's "not using it for project-specific stuff."
"Current status, schedule, assignment are on Web for ease of multisite and management viewing" at a Florida technical director's workplace.
Also from Florida, "E-mail, e-mail, e-mail." It's not clear if the writer was praising or swearing.
"Mail and multimedia cut travel and reduce mistakes," says another.
"Here in Podunk, Iowa," writes one ironic (and non-Midwest) engineer, "the 'library' in town is useless. We have no access to most suppliers or reps. So now that we have finally got Internet access I can get data from suppliers."
Indeed, at one Pennsylvania section head's workplace, "the library of data books has been eliminated. All data sheets are retrieved from manufacturers' Web pages."
While there was lots of favorable comment about the use of the Internet in engineering design, not everyone has embraced it.
"I have used the Internet occasionally in my work," says a South Dakota design engineer. "But I don't feel it has made a significant change in my work yet."
"A source of misinformation," another maintains.
"Too much information for most users without good filtering tools and processes," a Massachusetts section head writes.
IP cores
One new tool in the engineer's quiver is the IP core, a predesigned, preverified block that's intended for design reuse in a system-on-chip environment. A soft core consists of logic only, in the form of register-transfer level, Verilog or VHDL code plus test benches. A hard core includes a full physical layout. Examples of commonly used cores include DSP functions, PCI and USB interfaces and microprocessors such as ARM or M-Core.
This is still an emerging technology: just 24 percent of the 681 respondents report that they incorporate IP cores into their designs. The most frequent users of IP cores are technical directors (41 percent); vice presidents of engineering (46 percent) and software managers (37 percent). Lower down the ladder, only 12 percent of the senior engineers report any IP involvement. This is plausible because managers must always keep an eye on the budget. So if acquiring an IP core for a certain function will allow his team to not reinvent the wheel and move on to system matters, then it can be "a GIANT time saver, above all else," according to a Texas design engineer.
IP Cores: An emerging technology
(Figure)
"Significant savings in both time and effort," seconds an executive user of DSP, logic and specialty cores.
"It shortens the design cycle and enables the company to produce end products quickly," adds another user.
One California technical director who uses memory, logic and specialty cores confirms that they have saved time and effort. "We have a robust on-chip architecture that allows fairly easy reuse."
Of the responding users, 64 percent employ logic cores, 55 percent employ memory cores and 36 percent design with DSP cores. Another 43 percent tap IP specialty cores.
A nonuser writes that he's willing to jump on the IP bandwagon "if management can get past their issues with licensing."
One principal engineer has some doubts. The IP cores might save money and time, "but the end result is less skills, less innovation and more reliance on outside."
They're "saving company effort in terms of design, but time and money is now spent troubleshooting and interfacing with" the cores, according to a California designer.
A project engineer concurs with the designer, adding, "The overhead of communications and transfer and negotiation somewhat offsets" the cost and time savings IP cores allow.
"It's a tradeoff with time," says a consultant. "We spend a lot of effort getting further specs on specialty cores."
A Midwest design engineer offers this observation: "If the IP provider is part of the design team (i.e., DSP), it is well worth it. If the IP provider simply sells a piece of code, the time involvement is about the same as if it was developed internally."
A California department head sums up his feelings about using an IP specialty core succinctly: "So far it's been a fiasco."
Software: A Java future
The software tool box carried by the survey participants contains the languages and operating systems that one would expect. Multiple choices were allowed.
- DOS (61 percent)
- Unix (53 percent)
- Windows 95 (73 percent)
- Windows NT (51 percent)
- C and C++ (58 percent)
- Web browser software (61 percent)
The pattern is much the same in Singapore, Korea, Hong Kong and Taiwan. Heavy use of Windows 95 (83 percent) and DOS (56 percent), less use of Unix (27 percent) and C and C++ (28 percent). Japan follows a similar pattern.
U.S. respondents are eager to climb on the Java bandwagon. A mere 10 percent use it today; but 43 percent intend to employ Java in the next two years. Among Asian and Japanese respondents, only a single-digit percentage utilize Java today, but usage should roughly triple in the next two years.
The Windows NT express has already left the station with half the Americans who replied and that should climb to 60 percent in the next two years. Windows 98 and Linux were not included in the survey, but many may be expected to incorporate those systems by 2000.
The losers in the United States over the next two years will be DOS, Windows 3.1 and Macintosh. And Ada is almost off the radar with a 3.7 percent response, even though 16 percent of the respondents work with or come from the defense-electronics industry.
EE Times readers are primarily design and development engineers and managers and the publication is often perceived as a hardware newspaper. But the reality is that 22 percent of those who returned survey forms describe their work as primarily software and 35 percent say they work in both hardware and software. Plainly, in these days of IP cores and systems-on-chip, it's difficult to separate EEs into an either/or category.
Project status: late
One sharp example of the changes in the engineering workplace can be found in the response to the question about the length of the design cycle from specification to production.
A rather astonishing 37 percent of the responding readers of Nikkei Electronics Asia, 26 percent of those from Nikkei Electronics Japan and 19 percent of the Americans say their design cycle is less than six months. The pressure is on, they say.
Other typical deadlines are 6 to 8 months for 16 percent of U.S. engineers, 21 percent of the Nikkei Electronics Asia respondents and 18.5 percent of the Nikkei Electronics Japan group. Using 9-12 months are 23.5 percent of the Americans, 27.5 percent of those from Japan and 15 percent of the rest of the Asians.
Nearly three-quarters of those from Nikkei Electronics Asia, two-thirds of those from Nikkei Electronics Japan and more than half the Americans find themselves scrambling to wrap up projects in less than a year.
The product cycle, says one Texas technical director, "is frequently too time-to-market driven!" His last project lasted less than a year.
Project delays
Not everyone is running on a treadmill. Nine percent work on product cycles of more than two years. Typically, those tend to be defense projects.
So with this high-speed development train, it's no wonder that so many are late in arriving at their destination. More than 80 percent of responding U.S. engineers experienced delays in completing their projects this year. In past surveys, respondents have blamed technical snafus, management interference and on-the-fly changes in the specs for the tardiness.
The story is the same for the Nikkei Electronics Asia and Japan groups. Two-thirds of the engineers in Korea, Singapore, Taiwan and Hong Kong reported delays in their projects in the past 12 months; it's even worse in Japan, where 85 percent reported missed deadlines.
The "Career" chapter explores another reason managers think their projects have fallen behind: lack of technical personnel. More than half blame a lack of engineers for their failure to make deadlines. Engineers, on the other hand, reject that notion. As one respondent sees it, the real problem is upper management's setting "unrealistic deadlines."
Teams common in the U.S.
More than three-quarters (78 percent) of the U.S. respondents are assigned to a team, most with fewer than 20 other technical persons. About 35 percent have been part of a multifunctional team with marketers, purchasers or other nontechnical personnel.
And for the most part they believe the teaming approach works. It:
- saved money (49 percent);
- saved time (68.5 percent);
- shortened the design cycle (69 percent);
- reduced bugs (58 percent);
- improved the quality of the product (71 percent);
- and increased customer satisfaction (50 percent).
Similarly, most Nikkei Electronics Asia respondents are on a team, generally with five persons or fewer. Like the Americans, the team is likely to comprise all engineers.
However, one sharp difference emerges: Only two of those respondents reported being on a multicompany team; apparently there is less cross-fertilization going on in Taiwan, Singapore, Korea and Hong Kong.
Six out of 10 U.S. respondents and seven out of 10 Nikkei Electronics Asia counterparts have headed teams (the question was not asked of Japanese in the survey). Obviously, the managers are more likely to have done so than the engineers. Eighty percent of U.S. managers reported leading a team at some point.
Here are some of the most difficult challenges that respondents from both sides of the Pacific encountered in leading teams.
- A Colorado physicist: "Getting several companies and three universities in four countries headed in the same direction while maintaining budget and schedule."
- An Idaho engineer: "Allocation of risk and responsibilities. No shortage of people seeking credit, but a dearth of those willing to take responsibility."
- An Asian manager: "Matching the internal skill sets to customers' projects."
- A California project manager: "Getting all players to buy in to the team goal."
The common theme is people problems. Engineers can be ill-prepared to take on management tasks without some innate understanding of how to "gather up a herd of cats," as one team leader puts it.
"People are members of a team, but have other responsibilities as well, so it is hard to pull it all together for a specific time line," says a reader.
"People present problems that are much tougher to solve than those that electrons present," says one U.S. engineer. "People are composed of too many variables-we can't model them and predict their behavior!"
An Asian team leader was challenged "to make everybody feel that he is part of the team and is indispensable for the team's collective success."
One program manager loved the "excitement" that his multicompany team created within the company. But his toughest challenge was "eliminating non-believers or [those practicing] passive resistance. Some do not share vision and refuse to contribute or help form an alternate approach."
Another common problem that engineers and managers run into is ensuring team members proper recognition-and reward. A technical director says he found his toughest job came during "negotiations with human resources/personnel departments to properly compensate team members for their contributions." Those team members were lucky to have a high-ranking executive pushing the other departments.
Dealing with other areas of the company can be daunting. Another technical director says his biggest challenge was "convincing non-technical bosses what needed to be done and how the work was supposed to be done."
Incongruously, one manager says he found himself taking charge of a product-development team "in a technical area in which I did not have hands-on expertise or proficiency." Nevertheless, he says, he subscribes to the idea that teamwork saves money and time, reduces bugs and boosts quality.
Another reader says he battled "feature creep," the tendency for a team to add feature after feature to a straightforward product. He says he had to keep the team's eye on the end cost of the product.
How will your job change?
There seems little doubt that the engineering workplace will change over the next decade. Ten years ago, only a handful of government insiders were tapping the Internet, IP cores did not exist, fewer people were on teams (at least in the United States), and the idea of a six-month project deadline, from specification to the production stage, was extreme.
So what will engineering jobs look like in 2010?
Some offered personal views. "Hopefully, I will be in engineering management," says a Texas design engineer. "I expect to have led several engineering teams." Then he adds: "I will spend only two days at work and telecommute three."
That's not wishful thinking, at least according to a Wisconsin engineer. "High-speed networks will allow most high-tech employees to work from home or elsewhere," he predicts.
"I'd like to telecommute, work funny hours, find an honored place as a master craftsman and mentor-teacher," says a Washington, D.C., design engineer. But he thinks the reality will be harsher. "Probably I'll get too old, hurt and tired to keep up with technology and go teach high school."
A California technical director expects "More management, less hands-on technical participation, more strategy."
Career aspirations
What do you want to be when you grow up? Relatively few of the responding engineers and managers are eyeing the top rung of the professional ladder. Only 13 percent aspire to be president or chief executive officer, with those closest to that position the hungriest: 41 percent of the vice presidents of engineering want to be president vs. only 8 percent of the senior engineers.
Most respondents see themselves as entrepreneur (26 percent), chief technical officer (26 percent) or senior engineer (25 percent). Another 10 percent see themselves as consultants.
Nikkei Electronics Asia respondents are more ambitious but less entrepreneurial. About 31 percent seek to be president or chief executive officer of a company, but only 17 percent see themselves as entrepreneurs.
Return to 1998 Salary & Opinion Survey
RSS
