Intel's Lessons for Toyota

Most comments on the Toyota crisis focus on the company's reputation for superior, reliable vehicles and its nemesis in sudden acceleration incidents in many models, and in regenerative braking problems of the Prius.

Looking more carefully on the glowing press the company received until very recently, it turns out that the quality and safety of the product wasn't what intrigued many business writers. It was the system of improving the process, of which reducing defects was only a part. Two paragraphs from a typical feature, in Fast Company (2006), show how even admirers acknowledged the weaknesses of the Toyota system:

Lean manufacturing and continuous improvement have been around for more than a quarter-century. But the incessant, almost mindless repetition of those phrases camouflages the real power behind the ideas. Continuous improvement is tectonic. By constantly questioning how you do things, by constantly tweaking, you don't outflank your competition next quarter. You outflank them next decade.

Toyota is far from infallible, of course. In the past two years, recalls for quality and safety problems have spiked dramatically -- evidence of the strain that rapid growth puts on even the best systems [emphasis added]. But those quality issues have seized the attention of Toyota's senior management. In the larger arena, when the strategy isn't to build cars but to build cars better, you create perpetual competitive advantage.

Management evidently was not seized firmly enough. There's hope for Toyota--but only if the public is convinced that it is learning from the disaster and shifting from lowering costs while maintaining quality to enhancing quality without increasing costs. It won't be easy. The nature of software has made reliability a problem, even for premium-priced European manufacturers with strong work ethic traditions of their own.

The most optimistic precedent is Intel's response to the Pentium floating point bug of 1994, when some chips yielded erroneous results, inspiring jokes like:

Q: How many Pentium designers does it take to screw in a light bulb?
A: 1.99995827903, but that's close enough for nontechnical people.

At Intel, quality is job 0.999999998.

Q. What do you get when you cross a Pentium PC with a research grant?
A. A mad scientist. 

These are taken from a mathematics column by Ivars Peterson (1997) illustrating other software pitfalls. Of course, there were no fatalities associated with the flawed Pentium chip, but Intel, like Toyota today, was mocked for what consumers thought an inadequate response from a dominant brand. The chip maker recovered and even became the foundation of the new Mac operating system. The shock was probably good for it.

The mathematician who publicly revealed the Pentium flaw, Thomas R. Nicely, has an excellent summary of the episode and its lessons, which now speak to the auto industry, too, since one automotive information technology expert has called today's cars "30 or more computers on wheels":

Computations which are mission critical, which might affect someone's life or well being, should be carried out in two entirely different ways, with the results checked against each other. While this still will not guarantee absolute reliability, it would represent a major advance.

Have you forgotten those Intel jokes? Maybe you're too young to have heard them. Either way, that's potentially great news for the auto maker.