First Edition Dated September, 2000 24
If a key characteristic is not readily measurable within the production setting
(e.g., mean time to failure, leakage or vibration test results), the supplier
should identify lower level key characteristics that influence that key
characteristic. When measured and controlled, the lower level key
characteristics should ensure that the top-level key characteristic meets
Boeing requirements.
In addition, the supplier may find it valuable to flow high level key
characteristics down to lower level key characteristics in order to reduce
variation or improve quality or producibility in both the lower level processes
and the high level key characteristic.
The information in the AQS Control Plan should also serve as the basis for a
process database. A major goal of this database is the compilation of process
knowledge that can be used to conduct quality planning in advance of
production and to leverage the results. This knowledge will allow
manufacturers to build products right the first time by matching part
characteristics to machines or production processes. Design engineering can
also use process capability data to create designs that are more robust to
manufacturing variation.
Section 4.2.2 Variation control and reduction activities must
be performed on identified key characteristics. The approved
variation control method for a key characteristic is statistical
process control. Methods other than statistical process
control may be used, but measurable evidence must verify
that the process remains statistically stable and capable with
a Cpk exceeding 1.33. Other comparable process capability
index measures may be used.
When variation is a problem, the root causes of the variation need to be
identified and removed. Identification of key characteristics or key process
parameters is the first step in variation reduction activities. When key
characteristics are identified, some method of variation control must be
applied to them.
Under most circumstances, a statistical control chart will be the most effective
method to monitor and control variation. However, innovative use of tool
designs, product designs, process design and mistake proofing can also help
control variation. The goal is to reduce the variation of the key characteristic
so as to reduce the costs associated with that variation. The method of
variation control must be documented in the AQS Control Plan.
