The Pre-control Technique
Pre-control is a control charting methodology that uses specification limits instead of statistically-derived control limits to determine process capability over time. Pre-control charting is useful in initial process setup to get a rough idea of process capability. Pre-control charting does not use continuous data found upstream in the process which is more in alignment with prevention thinking.
An easy method of controlling the process average is known as “pre-control.” Pre-control was developed in 1954 by a group of consultants (including Dorin Shainin) in an attempt to replace the control chart. Pre-control is most successful with processes which are inherently stable and not subject to rapid process drifts once they are set up. Pre-control can act both as a guide in setting process aim and monitoring the continuing process.
The idea behind pre-control is to divide the total tolerance into zones. The two boundaries within the tolerance are called pre-control lines. The location of these lines is halfway between the center of the specification and specification limits. It can be shown that 86%of the parts will be inside the P-C lines with 7% in each of the outer sections, if the process is normally distributed and Cpk= 1. Usually the process will occupy much less of the tolerance range, so this extreme case will not apply.
The chance that two parts in a row will fall outside either P-C line is 1/7 times 1/7, or 1/49. This means that only once in every 49 pieces can we expect to get two pieces in a row outside the P-C lines just due to chance. There is a much greater chance (48/49) that the process has shifted. It is advisable, therefore, to reset the process to the center. It is equally unlikely that one piece will be outside one P-C line and the next outside the other P-C line. This is a definite indication that a special factor has widened the variation and action must be taken to find that special cause before continuing.
. Set-up: The job is OK to run if five pieces in a row are inside the target .
. Running: Sample two consecutive pieces
. If the first piece is within target, run (don’t measure the second piece)
. If the first piece is not within target, check the second piece
. If the second piece is within target, continue to run
. If both pieces are out of target, adjust the process, go back to set up
. Any time a reading is out-of-specification, stop and adjust
The ideal frequency of sampling is 25 checks until a reset is required. Sampling can be relaxed if the process does not need adjustment in greater than 25 checks. Sampling must be increased if the opposite is true. To make pre-control even easier to use, gauges for the target area may be painted green. Yellow is used for the outer zones and red for out-of-specification.
The advantages of pre-control include:
. Shifts in process centering or increases in process spread can be detected
. The percentage of non-conforming product will not exceed a pre-determined level
. No recording, calculating or plotting is required
. Attribute or visual characteristics can be used
. Can serve as a set-up plan for short production runs, often found in job shops
. The specification tolerance is used directly
. Very simple instructions are needed for operators
The disadvantages of pre-control include:
. There is no permanent paper record of adjustments
. Subtle changes in process capability cannot be calculated
. It will not work for an unstable process
. It will not work effectively if the process spread is greater than the tolerance