The Two Primary Causes of Product Deficiencies
Many consumers have experienced problems with a product, whether it’s a furnace, air conditioning system, appliance, automobile, electronic device, food, medical device, etc. Some products fail prematurely, and as a result, manufacturers spend billions of dollars a year on warranty and product litigation costs. In fact, many companies, such as large automakers spend several billion dollars each year fixing products that fail within their warranty periods. The direct costs of premature failures can drastically reduce the profitability of manufacturers. Furthermore, the indirect costs due to a tarnished reputation and lost market share also impair companies significantly.
There are three typical causes of premature product failures, but two are predominant and are directly controlled by manufacturers (if they care to pay attention). The three are:
- Design Flaws (incorrect specifications on materials, dimensions, and/or processing parameters)
- Excessive Variation (in manufacturing operations)
- Customer Abuse
The first two causes on the list are the predominant causes of financial losses, and the manufacturer directly determines their adequacy.
Many times, designers of products are not sure how to specify material properties (characteristics such as hardness, yield strength, carbon content, etc.) or dimensional features (such as gaps, thicknesses, concentricity, etc.). Often, they have not performed adequate testing in order to determine proper limits on these features. Furthermore, they do not predict how the product will work in a variety of different environments.
Here, industrial statisticians like those at Integral Concepts can be extremely helpful in minimizing the number of tests and type of testing that should be performed. Industrial statisticians can also develop mathematical models which predict HOW the products will perform given numerous possible conditions of the material, dimensional properties, and many environments. Unfortunately, many manufacturers do not receive input from industrial statisticians or mathematicians.
Imagine that a particular brand of AA alkaline batteries were absolutely identical (which is sadly impossible). If “identical” batteries were placed into the same device and used, they should fail at exactly the same time/usage.
The reality is that batteries placed in the same device do not fail at the same time/usage. Some last longer than others, and the reason is VARIATION among the batteries. For example, some of the batteries may have more chemicals than others, due to variation in the chemical dispensing operations. The batteries may have different ratios of chemicals—again from variation in the manufacturing of the batteries.
Industrial statisticians working in manufacturing environments identify detrimental sources of variation that affect product performance and work to minimize that variation. As a result, product performance is more consistent in the marketplace/application, so users of the product share similar experiences, and warranty costs are minimized.
The most important statistic to emphasize in manufacturing environments is called the “standard deviation,” and it measures the amount of variation in important characteristics. Many world-class manufacturers work to minimize the standard deviation in key characteristics, while other manufacturers ignore variation and deal with the steep financial, legal, and market ramifications.