February 2, 1998

Who said anything about spare parts?

Bill Schweber, Technical Editor

As we crank out the next faster, smaller, cheaper widget, keep in mind that the earlier versions are not obsolete from a user's perspective.

I recently needed a new puck for my 10-year-old "knock-hockey" set. I called the maker, whose name and phone number were on the set, and the person who an-swered said, "No problem. We'll send you six new pucks for $5, postage included, just give us a credit-card number." It was that simple, and replacement pucks arrived a few days later.

In our rush to put new products on the market, it's easy to overlook the need to sustain existing products or previous generations. Contrary to what we would like to think as we crank out the next fast-
er, smaller, cheaper widget, keep in mind that the earlier versions are not obsolete from a user's perspective. Sure, the new ones are much better, but that's not the point.

In many cases, the existing system is quite satisfactory and sufficient for the person or or-ganization that bought it. When it fails, there's a strong user-comfort factor in having it repaired. Some of this user comfort relates to cost, but a lot of it results from the user's familiarity with and confidence in the product, as well as its functionality. This situation is especially true in critical industrial and commercial applications. Sometimes, that next generation has subtle performance changes, added or removed features, or modified operating procedures that users can't--and shouldn't--be forced to accept. Certainly, today's reality makes it difficult for users to logically justify repair for many items. Repair costs, such as shipping and labor, are often comparable with complete replacement cost.

If that's the case, why should you bother worrying about spare parts for three or four years from now? The answer is that what's easy and makes sense for you may not be the right thing for your customer. That's why you need to look at your product's design and use, along with its bill of materials and decide how you will ensure critical spare-part availability.

When manufacturers built systems primarily from standard-function analog and digital ICs, discrete devices, and passive components, part availability was less of a problem than it is today. Even if you couldn't get the exact replacement, you could obtain a functionally similar equivalent. This situation has changed with the increased use of ASICs, custom and semicustom devices, embedded code, and even passives with extremely sophisticated design and performance--and there's no going back to the old days, either.

What should you do? Look at your design. Which of its parts are unique? Which parts have certain nonobvious or secondary characteristics that you rely upon? Al-though today's electronics are amazingly reliable, certain parts of a system are more likely to fail because of misapplication or heat, such as I/O-port-related components, connectors, and power-supply devices. At least be sure to thoroughly document your bill of materials' subtleties, so that users can find alternative sources in the future.

If you think you can't do this task or that it is too much work, think again. Even the often-maligned auto industry, which releases new models every year, has successfully dealt with the problem. You can get a replacement windshield for a 10-year-old car within 24 hours in most cases, for example. Parts for older cars are usually available directly from the manufacturer via the dealer or from secondary vendors that specialize in after-market parts. Because the electronics industry has relatively few of such follow-up vendors, you have to take responsibility for making sure that your source is going to be there when you need it, even after you view your product as obsolete. 

EDN Access | Feedback | Table of Contents |