Page 122 - Six Sigma for electronics design and manufacturing

91
Six Sigma and Manufacturing Control Systems
3.3.7 Use of control charts in factories that are
approaching six sigma
The C chart is the most widely used chart in factories that are ap-
proaching six sigma. Since the defect rates are very low, binomial-
based control charts would require a very large sample, and hence are
impractical to use. For six sigma quality, a defect rate of 3.4 PPM
would result in a nP chart with a centerline probability 0.0000034.
Such a chart would require a very large sample to determine if the
process indeed has gone out of control.
Using C charts with well-defined areas of opportunity, such as de-
fects per shift or defects per 10,000 units, can be effective for monitor-
ing quality control in production. In some factories, the discussion has
shifted to the number of possibilities of defects, or the number of op-
portunities. The electronics industry has defined a new C chart met-
ric, the DPMO (defects per million opportunities) chart. A discussion
of DPMO concepts and calculations is given in Chapter 4.
A more realistic way to achieve quality control in factories that
approach six sigma is to closely couple the total defect reporting to the
continuous quality improvement team. The low defect rate of six sig-
ma manufacturing operation would produce a small number of total
defects per day, even in a large factory. For example if we assume
that a factory produces 5000 PCBs per day, and each PCB requires
2000 operations, that is a total defects opportunity of 10 million oper-
ations per day. For the six sigma defect rate of 3.4 defects per million,
the total expected defects is 34. The management of the factory can
review these defects individually each day, then decide what correc-
tive action is needed, whether immediate, short, or long term. They
can use the tools of TQM to monitor, organize, and rank defects and
initiate a corrective action plan to reduce them further.
3.4 Using TQM Techniques to Maintain Six Sigma
Quality in Manufacturing
When factories approach six sigma quality, the need for control charts
with their sampling-based methods is reduced. The quality team can
review all of the defects that occurred each day in production, using
the TQM tools to effectively manage the corrective action process.
Table 3.4 shows a list of TQM tools grouped into three major areas
according to their use: including tools for data analysis and display of
problems, then tools for generating ideas and information about a
likely solution, and then tools for decision making and consensus for
the TQM team to resolve the problems. In the example of the factory
in the last section that generates 34 defects per day the procedure for
corrective action could be as follows: