Building a better mousetrap can be key to economic success. But sometimes production techniques go awry, and it’s the manufacturer who is caught with defective products that cannot be shipped.

In today’s complex manufacturing environment, these problems can be difficult to find and correct. A technique that I have used successfully on products from roast beef to machine guns has helped more than 100 northern New England companies to build better products, save millions of dollars a year, increase production and protect jobs in local communities.

Consider the producer of high- quality roast beef, for example, where 1.5 tons of meat are cooked simultaneously in large industrial ovens. In our own kitchens, we may see the unappetizing result of ovens that have hot spots. At a food production company, such problems can be disastrous for profits.

No matter what the employees did – continually reposition the carts of meat, vary cooking times and temperatures – the meat was cooking unevenly, and there was too much waste.

I normally work with skilled teams from each company, so it was not necessary for me to be a commercial roast beef cook. I sat down with the cooks and the company’s engineers and helped them to identify all the possible factors that could cause the problem.

Then we designed a set of tests. For 11 possible causes of the problem, there were more than 2,000 possible combinations to test, far more than would be practical to actually run. It would be too expensive and take too much time. Most of all, it is unnecessary.

Instead, I used statistical methods to design a much shorter testing plan with only eight combinations of factors that would provide significant information about cooking variations. When we ran those eight tests, we found that one step in the cooking process was the culprit. By correcting the method of loading the oven, we reduced the cooking variation and generated more uniform high-quality roast beef. As a result, we saved over $120,000 in annual production costs.

This technique is called DOX, or Design of Statistical Experiments. It has worked equally well on fiber optics for surgical lasers, Velcro, anti-lock brake seals, guns, semiconductor production machinery, paper manufacturing processes and stainless steel tubing for aircraft.

More than 100 companies in Maine, New Hampshire, Massachusetts and Canada have taken this smart manufacturing approach to solving their production problems and thereby increasing profits. In fact, this program has won two national awards for innovation and effectiveness.

Sometimes the solutions defy common sense. For example, sand- blasting metal plates had always been considered a necessary step in producing brake seals. The company swore by it. However, the tests showed that sand blasting actually produced a worse seal. Eliminating that step saved money and reduced the failure rate to zero.

In another case, a manufacturer of glassware for medical laboratories double washed their products to ensure cleanliness. The DOX tests showed, however, that slides washed only once were actually cleaner. Moreover, the company was able to identify problems with a specific glass manufacturing machine from a supplier in Europe.

Companies have learned important lessons, some of which are like ones we learned in kindergarten, as author Robert Fulghum would say. Cooperation and teamwork are critical. Maintaining a stable work environment is important for consistent quality improvement. Well-planned, methodical experiments generally work better than a seat-of-the-pants approach. Success de-pends on the cooperation and commitment of both management and the people on the shop floor.

The DOX program is a true collaboration between industry and academia. Companies learn not just to solve a particular problem but to apply a scientific method to their production processes. Once learned, the techniques can be used repeatedly to improve performance.

At the University of Maine, the Department of Industrial Cooperation is our gateway to working with Maine companies. In turn, our faculty members stay current in manufacturing practices and keep the university’s engineering technology courses effective and relevant. Our students have also been involved and have obtained professional experience as interns.

But most important, we have been able to contribute to continuing economic success, through the use of these smart manufacturing techniques to build profits and protect jobs in Maine.

Jill Schoof is an associate professor in the Electrical Engineering Technology Program at the University of Maine.