Reducing Warranty Claims
by Reliability Improvement
Reliability is dened as the probability that a device will perform its
intended function during the specied period of time under the specied
conditions. Mathematically, this may be expressed as a percent probabil-
ity, or as the expected failure quantity rate. It may also be expressed as
average elapsed time to failure or time between failures.
ere are a few key elements of this denition. Performance is the exhi-
bition of the proper functioning of the product. It is the specied function.
is function may be specied by whoever has design authority, either the
manufacturer or the customer, or the end user. Performing the intended
function is taken to mean operating in the prescribed manner without
failure. However, even if no individual part of the product fails, but the
product as a whole does not do what it was intended to do, then it is still
considered to be a failure.
Reliability applies to a specied period of time. In practical terms, this
means that the product has to operate without failure until that specied
time period has been completed. Units other than time, such as number of
operations or miles of travel, may sometimes be used instead.
Reliability requires satisfactory operation within certain conditions
stated in the performance specication or the reliability specication.
ese are oen explicitly dened. is is necessary because it is impos-
sible to design a product for any and all conditions. A commercial aircra
has to operate under conditions of temperature, wind speed, pressure,
vibration, and altitude that are far dierent from the conditions in which
a child’s toy must operate. e operating environment must be considered
during design of the product. Testing of the product to verify and vali-
date the design must duplicate or simulate these conditions as closely as
88 • Warranty Claims Reduction
possible. e more realistic they are, the more valid the test. Invalid test-
ing is a common cause of inaccurate reliability predictions.
In the case of consumer items, it is the manufacturer that species the
conditions under which a product must function. It is the consumer’s
responsibility to decide if the product is suitable for his or her particu-
lar intention. But, with custom- made products, it is the customer that
determines the requirements, and it is the manufacturer’s responsibility
to manufacture the product so that it meets the customer’s requirements.
is is especially true when the manufacturer has design authority, or if
the performance conditions are previously agreed upon between the man-
ufacturer and the customer.
For products that are not consumer items, the reliability requirements
are explicitly specied. is is usually done in the contract statement of
work, or in a product performance specication. e level of detail may
depend on how much control the customer wishes to have over the con-
tractor’s activities. It can also depend on the severity and variety of envi-
ronments within which the product must function.
Determining reliability involves a series of tasks, including component
research, various analyses, planning, testing, and failure reporting, as well
as communication with design and manufacturing engineers. An item
that is critical to safety may require an actual formal failure reporting and
review process throughout its development, whereas a noncritical system
may rely only on nal reports and published reliability data.
While it is true that reliability engineering is normally done during
product development prior to launching the design into production, there
is also a legitimate role for reliability engineering that occurs aer produc-
tion has been going on for a while and products are being returned for
warranty claims. is dual role of reliability activities makes reliability
doubly important to warranty claims reduction.
In its rst role reliability engineering occurs during the product’s initial
development. It predicts failure rates and anticipates causes of early fail-
ures, thereby giving design engineering and manufacturing insight into
what needs improvement. e improvements themselves are the result
of successful corrective and preventive actions determined necessary by
the reliability test results. ese actions are implemented before going
into the full production mode and will reduce or eliminate early failures,
thereby reducing the number and frequency of warranty repairs and all
of their associated costs.