September 12, 1997

Whatever happened to second sources?

In today's fast product cycles, there's simply no time for the luxury of identifying real or potential second sources.

A recent article in the Wall Street Journal (reference) detailed Toyota's response to a fire at one of its supplier's factories. This factory produces nearly all of Toyota's brake valves, and such a fire could have shut down the company's car production in Japan. Toyota's entire supplier community--including primary, second-level, and even third-level suppliers--rallied to produce new tooling and parts in five days of nonstop, all-out effort. None of the suppliers asked for any payment, despite the extraordinary dislocation and costs involved. These suppliers knew that their long-term success depended on Toyota's shipping cars, and they would reap their rewards after the crisis.

In electronic design, many components, similar to Toyota's brake valves, are also unique. This fact ex-plains why you often use such components: for the competitive advantage they bring. Some components may have second sources, such as form-, fit-, or functionally compatible ones, or alternative sources, which are similar enough to the original devices that they could work but would require some redesign. However, these components are the exceptions in today's designs.

The brake valve has an inherent advantage over most electronic components. Unlike electronic components that have a longer initial production cycle and are critically dependent on the vendor's fabrication technology, Toyota's parts are machined items. Even small local shops could help with fabrication once they had the drawings and tooling. Compare this situation with fabricating a leading-edge VLSI device and the steps it involves, even if you have full documentation.

Not long ago, a good designer was expected to identify second and alternative sources for nearly all the devices in the design. To some extent, this task was easier to do when analog components and basic digital logic dominated designs, because these devices tend to have more form- and fit-compatible alternatives, which differ primarily in performance subtleties. The migration from small-logic-function ICs to ASICs and large-scale ICs now means that nearly equivalent devices are nonexistent.

What has really changed, though, is that the competitive pressures on designers to quickly get a product with superior performance to market largely negate the opportunity to use second-source parts. You must choose the devices that give you that competitive edge, even if they are available from only one vendor.

How does this situation affect your design strategy? Acknowledge that you're betting your product's viability on single-source components. Take extra time to qualify vendors by investigating issues such as: Do they have a track record? Can they deliver? How can they ensure delivery? If there is a crisis, where are you on their priority list? How far would they go for you if they had a catastrophe or if they lost the process recipe? Are you dependent on the subtleties of that supplier's process?

Also, ask yourself what you can do for insurance. Do you have documentation that would help you line up another supplier if needed? Have you at least investigated other potential sources with suitable fabrication processes, just in case?

In today's fast product cycles, there's simply no time for the luxury of identifying real or potential second sources. Sole-sourcing is a chance you have to take, and the best you can do is assess the track record and credibility of your vendors--although, as the Toyota case points out, even a stellar record is no assurance that calamities can't halt availability. Give sole-sourcing your best shot, but recognize today's design reality.

References

1. "Toyota's fast rebound after fire at supplier shows why it is tough," Wall Street Journal, May 8, 1997, pg 1.

EDN Access | Feedback | Table of Contents |