Profiles: Learning From Segway: Innovation in Action

Columns From: 1/1/2006 Automotive Design & Production, , Editor-in-Chief from Gardner Business Media, Inc.

Here’s how J. Douglas Field, vp of Design and Engineering and Chief Technology Officer of Segway, sees the challenges and opportunities of product development. Know that he spent some time working in auto, so his ideas are not totally on the bleeding edge of development.

It’s been a while since J. Douglas Field worked at Ford Motor Co. as a vehicle development engineer and team leader, where his areas of work included ride, handling and NVH. A graduate of Purdue (BS in mechanical engineering) and MIT (part of the Leaders for Manufacturing Program, he earned a dual degree: a masters in mechanical engineering and an MBA from the Sloan School), Field recalls that he wanted to experience another industry, so he went to work on medical products at Johnson & Johnson Medical. He then met Dean Kamen, who, it should be noted, has developed a number of medical devices, including dialysis equipment and intravascular stents. “He told me about a project that he thought was my dream project,” Field says, adding, “And Dean is very compelling when he tells you things like that.” Field was hired by DEKA Research & Development Co. (; Manchester, NH) in 1996, where he worked on what was to become the iBot, which he describes as “a balancing, stair-climbing, robotic wheelchair-like device,” which is available from Independence Technology (, a Johnson & Johnson company. “While at DEKA, I fell in love with the idea of creating personal transportation off of the same technology,” and as Kamen went on to establish Segway (, Field found himself leading the development team for what became the Segway Human Transporter (see Panic In Detroit). Today field is vp of Design and Engineering and Chief Technology Officer for Segway.

GETTING THE RIGHT PEOPLE. Although the work of developing personal transportation devices is moderately distant from vehicle development, there are certainly some things that are being done at Segway that could be of immense value to those in the auto industry. For example, consider what it takes to be hired onto the development team. “To become a member of the team there’s a bar to jump over that’s not only based on how much you know”—which is pretty standard for most job interviews—“but on how well you operate when you don’t know what you’re doing.” He explains, “No matter how smart somebody is, you can always get them to a point where you stump them. The way people behave when they’re stumped is a better indicator of whether they’re the right person for the environment even more than how much they know.” Field says some people make up answers. They’re dangerous because they don’t let on when they’re beyond their knowledge base. There are people who shut down. They may actually be useful late in program development. Finally, there are the people who get excited by the fact they’re learning something they didn’t know: “The sort of gems that flourish in the kind of environment where you’re doing something that hasn’t been done before.” There are about 30 people on the design engineering team at Segway, people who represent various engineering disciplines (from mechanical to software) and who range from academic PhDs to people who “grew up in casting plants.”

GETTING THE RIGHT VARIETY. This array of people is important in order to get fresh ideas, particularly since, Field explains, unlike other industries like automobiles and airplanes, light electric transportation doesn’t have a “dominant architecture,” so, “we want to make sure we have the freedom to explore really crazy ideas for some period of time.” One of the means by which these ideas are generated is through what they call “Frog Days” at Segway. “People are incentivized to make colossal failures, but to do it very quickly, and to explore crazy ideas with the endorsement of the company.” There are tight timelines, and failure doesn’t come along with what it is often partnered with: reprisal. However, this is not all about crazy ideas, but of developing salable, commercial products. Field says that as they move along in their development programs they “start to look more traditional.” But there’s a twist. During design reviews, it is not those who have the traditional authority who matter most, but those who have the knowledge, experience, and responsibility. “We have a language we call ‘Pigs, Chickens and Cows’ for our major projects.” It is about the level of commitment, as in the pig being the most committed to a ham-and-egg breakfast. They have the most at stake. The chickens are technical experts and advisors; the cows are consultants to the program, without a lot at stake. “You’ll hear people at meetings asking who is the ‘pig’ on this decision rather than saying who is the vice president or the manager that’s going to make the decision.” Knowledge trumps status.

GETTING THINGS DONE—FAST. Another thing the engineers are encouraged to do is prototype. Physically and frequently. “We iterate in hardware,” Field says. “An engineer always knows their P.O. will get approved to build something.” They build in foam. Stereolithography. Cast urethane. Machined parts. Once again, there is the issue that they are creating new things, products that aren’t aligned within a dominant architecture. Consequently, Field says that using CAD or doing virtual prototyping is not the appropriate approach. He explains, “In my experience, that works well when you have incremental design or very large programs that you’re spending.” He goes on to say that “virtual prototyping tools start to encourage designs that work in those tools rather than solve problems in the best way, so you end up with a hammer looking for nails.” In other words, to create something new it isn’t useful to have something that is pushing you down a path that has already been established. “The more you invest in infrastructure that makes assumptions about what your product is, the more likely your product will turn out that way at the end.” You won’t end up with something different. Innovative. “What the tools do well is extrapolate. And extrapolation works well when the points aren’t too far from the data points you have. If you’re incrementally evolving a product line, it’s probably a very rational way to do things. But if you want to create really revolutionary products, I think it’s crazy.”

GETTING SOUL. “Products need stewardship,” Field says. “They need ownership. When responsibility for making a product great is distributed across organizations and people, products very quickly lose their soul.” They lose the essence of what can make them desirable and different. So what Field recommends for product development is to have small groups of people who have responsibility and authority. “The way Kelly Johnson ran the Skunk Works at Lockheed is a great example of how things need to be managed to insure both innovation and the basic throughput of products,” he notes (see Mach Speed Product Development: Inside the Skunk Works). He acknowledges, however, that the complexity of an automobile—both from the standpoints of the various components and systems that it comprises as well as the economic and legislative contexts in which it must perform—makes its development a difficult task. Still, he thinks that it might be useful to the automotive industry if there were more small groups, or what he calls “internal innovation centers” that would help companies be positioned to seize opportunities that may be presented by changes in technology such as hybrid powertrains and by-wire technologies. He points out, though, that these would not be “research” groups, but enabled organizations that would be changed with creating “demonstration products and prototypes that could really be considered for potential introduction.” 

GETTING GUIDANCE FROM DARWIN. Field cites Darwin’s Theory of Evolution as a guide to product development. He puts “product” in the place of Darwin’s “species.” He suggests that while there are three mechanisms involved in evolution, companies tend to concentrate on just one: “I think most big companies have survival of the fittest down. They kill off bad projects, and push projects to be better and better, to have lower and lower costs.” But that’s not wholly sufficient. The other two—mutation and the production of too many offspring for all to be sus-tained—are things he thinks there ought to be more attention paid to. “Mutation is not something you control, but that you allow and enable in an organization,” he says. Mutation is random. It is necessary for the advancement of products. Most companies try to stamp out the possibility of randomness. As for the third mechanism, the production of plenty of offspring, most of which don’t make it, he admits that it is “an unpleasant thing for people to think about,” as it could mean that perhaps 50% of projects or more should be canceled before they get to product. “If you have an environment where it is a career dead-end to have your project canceled, then every incentive in the company is going to line up around making a project successful in the survival of the fittest system, which means it’s not going to be that innovative.” People will do what they’ve pretty much always done. Field is not in favor of putting products out into the market that will fail, especially in industries like automotive where a product failure can put an entire company at risk. He’s talking about smaller-scale initiatives before production. “The fail-faster-to-succeed-sooner mentality is really, really important.” It’s demonstrated development—plenty of developments—prior to the market that can lead to impressive innovations, or not. “‘Do it right the first time’ only works if you’ve done it before. If you’re doing something that’s never been done before, that’s a preposterous statement. The systems and organizations put in place that thinks ‘do it right the first time’ is one based on incremental improvement. It’s hard to necessarily criticize a company like Ford or GM for doing that.” However, if a company like that really wants to be an innovative leader. . . .