Test: A value center.

By Glenn Woppman
President and CEO
ASSET InterTech

I call test a value center because that’s what it delivers. This value comes in a variety of forms. The first to come to mind is the most obvious: product quality. But from quality the company derives a whole raft of other benefits, such as satisfied and loyal customers, lower support costs, fewer product returns, less re-work…the list goes on and on. At one point, let’s say six months ago, the Toyota brand and quality were practically synonymous. Now, we’ve witnessed how quickly public trust can dissipate. Toyota has been forced to undertake expensive programs to recover the trust and admiration that had made it the envy of the industrial world.

Of course, what I’m saying is not new for the professionals of the test industry. It is part of our mantra. But there are certainly plenty of professional test engineers and technicians, and probably a vast number of quality-control personnel within Toyota. Maybe if the test function had been perceived as a value center rather than a cost center, Toyota might have fared better throughout this ordeal. But, it would be too easy and unfair of me, an outsider, to lay blame at the feet of Toyota’s test professionals or its management. What I can say, as someone intimately involved in the test and measurement industry, is that I believe test personnel must be able to demonstrate the value of test to management and convince them that test is an essential contributor to any company’s existence.

Making our case

I suppose that if someone were to ask me the question: Why test? I would have a knee-jerk reaction and talk a lot about cost reductions. I bet many test professionals would center this discussion on cost reductions and rightfully so. We have a lot of ammunition there, but this approach places test on the cost side of the ledger. We want test to be on the value side, because that’s what it generates. Test enhances the value of products by improving their quality. Higher quality products command higher prices, find more loyal users and the benefits go on and on.

So we have to talk about value, but filibustering on value won’t convince anyone unless test is indeed valuable. What I mean is: we, the test industry, must be committed to providing high-quality test results cost-effectively.

We’ve probably all heard this comment from time to time: “The simulation is enough. We don’t need to test.” That kind of comment reminds me of a quote from a famous statistician and quality expert, George E. P. Box:

“All models are wrong, but some are useful.”

I think Box was saying that all simulations are only approximations of reality. They’re not the real thing; by definition, they will always fall short. Simulation should never replace or be used as a substitute for good hard empirical evidence. Some simulations are useful because we can’t get our hands around the object or event being simulated. If we could, we could and should test it empirically. That brings me to another of our mantras: If you can’t measure it, you can’t improve it! Simulation can certainly be a helpful tool, but in the end it’s a facile substitute for empirical test measurements. After all, the value test has to offer is in the improvements it can cause in many areas.

I realize that those nagging day-to-day tradeoffs involving profit and loss almost always get in the way of ideals, but that doesn’t mean that those ideals are any less worthwhile to pursue. 

Test as a corporate linking strategy

Another valuable aspect of test that is becoming more real every day even though you won’t see it on any list of traditional test values like the one above is the potential it holds as a corporate linking strategy. This evolutionary concept could certainly become a game-changer. Let me explain.

Every company is made up of functional silos. The design department is in one place while manufacturing is in another part of the building or in a separate building in another city or country entirely. Pre-sales support is separate and distinct from post-sales support. One of the great challenges facing large and small companies is how to breach the boundaries that grow up quite naturally among functional groups. In its own way, test can help pull down some of these walls.

The day is fast approaching when tests, diagnostic routines, validation processes, debug programs – whatever you want to call them – they will all flow freely from the very beginning of product design right through to post-sales and in-the-field support. Instead of walls between functional silos, the lines between groups will look more like organic membranes and test will flow from one group to the next as if by osmosis.

As idealistic as this may sound, there are very sound financial reasons why companies should evolve in this direction and we, as test professionals, should encourage it. For one, it increases the value – There’s that word again! – of test tremendously. A test developed when a chip is being designed that can follow the chip to prototype debug in design, and then flow through to board-level manufacturing test and eventually end up in a support engineer’s arsenal of diagnostic tools – now that’s a valuable test!

Next we start working on the financial guys to add value centers to the ledger.