Please ensure Javascript is enabled for purposes of website accessibility
Powered by Zoomin Software. For more details please contactZoomin

AVEVA™ Manufacturing Execution System 2023 R2

Computing from standard values

  • Last UpdatedNov 04, 2024
  • 9 minute read

When the cl_source for a characteristic or for a QM specification that overrides the characteristic is configured to use the given standard mean and standard deviation (cl_source = 1), the given standard mean (std_avg) and the standard deviation (std_deviation) are used to calculate the control limits to validate the measurements recorded for a characteristic.

If the standard mean (std_avg) or the standard deviation (std_deviation) for a characteristic is not given, the mean or the standard deviation (estimate or population) is calculated from the data. The standard deviation is calculated based on the following:

  • The sigma estimate (sigma_est) configured for a characteristic.

  • The default chart type linked to the characteristic.

If the standard deviation is given, it is adjusted based on the chart type.

The mean and standard deviation are only used to calculate control limits for a characteristic. These values are not used for calculating statistical values such as Cp and CpK. The target value configured for a specification, that is associated with the characteristic is used as the standard average in the following conditions:

  • The given standard average (std_avg) contains a non null value.

  • The std_avg_is_target is set to True for the characteristic.

The following table describes how the given standard deviation, CL, LCL and UCL are adjusted based on the chart type. The examples given below have the following:

  • Standard mean set to 0.6 for variable-type characteristics

  • Standard deviation set to 0.8 for variable-type characteristics

  • Standard proportion set to 0.6 for binary-type characteristics

  • Standard defects per unit set to 10 for counted-type characteristics

The examples in the table below are only applicable to the final chart point that can be plotted in a chart for a characteristic. The control limits for the other chart points can be calculated using the formula for a chart type.

Chart Type

Formula

Comments

Example

ix

CL = Given Standard Mean

LCL = CL – (3 * Given Std. Dev.)

UCL = CL + (3 * Given Std. Dev.)

Since individual data is considered for this chart type, the sample size is always 1.

CL = 0.6

The given standard mean is 0.6, and the given standard deviation is 0.8.

LCL = 0.6 – (3 * 0.8) = –1.80

UCL = 0.6 + (3 * 0.8) = 3.00

mr (ix + mr)

CL = Given Std. Dev. * d2 [2]

LCL = CL – (3 * (Given Std. Dev. * d[2]))

UCL = CL + (3 * (Given Std. Dev. * d[2]))

if LCL < 0, then 0; otherwise LCL.

Since this is a moving range chart, the range is always between 2 measurements, hence the d2 factor is always for sample_size=2.

CL = 0.8 * 1.12838 = 0.90270

The given standard deviation is 0.8 and the d2 factor value for a sample size of 2 is 1.12838.

LCL = 0.90270 – (3 * (0.8 * 0.85250)) = – 1.14330

Since LCL is less than 0, LCL = 0

UCL = 0.90270 + (3 * (0.8 * 0.85250)) = 2.94871

The d3 factor value for a sample size of 2 is 0.85250.

mr
(ma + range)

CL = Given Std. Dev. * d[Moving Average Span]

LCL = CL – (3 * (Given Std. Dev. * d[Moving Average Span]))

UCL = CL + (3 * (Given Std. Dev. * d[Moving Average Span]))

if LCL < 0, then 0; otherwise LCL.

The moving average span configured for a characteristic is used to adjust the given standard deviation for calculating the control limits.

CL = 0.8 * 1.69257 = 1.35406

The given standard deviation is 0.8 and the d2 factor value for a sample size of 3 is 1.69257.

LCL = 1.35406 – (3 * (0.8 * 0.88837)) = – 0.77803

Since LCL is less than 0, LCL = 0

UCL = 1.35406 + (3 * (0.8 * 0.88837)) = 3.48614

The d3 factor value for a sample size of 3 is 0.88837.

xbar

CL = Given Standard Mean

ASTD =

LCL = Mean – (3 * ASTD)

UCL = Mean + (3 * ASTD) where ASTD is the adjusted standard deviation from the given standard deviation.

The sample size from a sample is used to adjust the given standard deviation for calculating the control limits (center line).

ASTD = = 0.4

The given standard deviation is 0.8 and the sample size obtained from this sample is 4.

CL =0.6 (Mean)

LCL = 0.6 – (3 * 0.4) = –0.600

UCL = 0.6 + (3 * 0.4) = 1.800

range

CL = Given Std. Dev. * d[Sample Size]

LCL = CL – (3 * (Given Std. Dev. * d[Sample Size]))

UCL = CL + (3 * (Given Std. Dev. * d[Sample Size]))

if LCL < 0, then 0; otherwise LCL.

The sample size from a sample is used to adjust the given standard deviation for calculating the control limits (center line).

CL = 0.8 * 2.05875

CL = 1.64700

The given standard deviation is 0.8 and the d2 factor value for a sample size of 4 is 2.05875.

LCL = 1.64700 – (3*(0.8*0.87981)) = –0.46454

Since LCL is less than 0, LCL = 0

UCL = 1.64700+(3*(0.8*0.87981)) = 3.75854

The d3 factor value for a sample size of 4 is 0.87981.

sigma
(xbar + sigma)

Sample Size = Sample Size of a sample

CL = Given Std. Dev. * C4 [Sample Size]

LCL = CL – (3 * (Given Std. Dev. * ))

UCL = CL + (3 * (Given Std. Dev. * ))

if LCL < 0, then 0; otherwise LCL.

The sample size from a sample is used when adjusting the center line, LCL and UCL.

CL = 0.8 * 0.92132 = 0.73705

The given standard deviation is 0.8 and the C4 factor value for a sample size of 4, obtained from this sample is 0.92132.

LCL = 0.73705 – (3 * (0.8 * )) = – 0.19609

Since LCL is less than 0, LCL = 0

UCL = 0.73705 + (3 * (0.8 * )) = 1.67020

ma

CL = Given Standard Mean

ASTD =

LCL = Mean – (3 * ASTD)

UCL = Mean + (3 * ASTD)

The adjusted standard deviation from the given standard deviation is ASTD.

The moving average span configured for a characteristic is used to adjust the given standard deviation for calculating the control limits (center line).

ASTD = = 0.46188

The given standard deviation is 0.8 and the moving average span configured for a characteristic is 3.

CL =0.6 (Mean)

LCL = 0.6 – (3 * 0.46188) = – 0.78564

UCL = 0.6 + (3 * 0.46188) = 1.98564

sigma
(ma + ms)

CL = Given Std. Dev. * C[Moving Average Span]

LCL = CL – (3 * (Given Std. Dev. * ))

UCL = CL + (3 * (Given Std. Dev. * ))

if LCL < 0, then 0; otherwise LCL.

The C4 value is based on the moving average span configured for a characteristic.

CL = 0.8 * 0.88623 = 0.70898

The given standard deviation is 0.8 and the C4 factor value for a moving average span of 3 configured for this characteristic is 0.88623.

LCL = 0.70898 – (3 * (0.8 * )) = – 0.40282

Since LCL is less than 0, LCL = 0

UCL = 0.70898 + (3 * (0.8 * )) = 1.82077

p

CL = Std. Mean (Given)

LCL = CL – (3 * )

UCL = CL + (3 * )

where n is the actual sample size from a sample, and Std. Mean (Given) is the given mean value.

If LCL < 0 then LCL = 0; otherwise LCL.

If UCL > 1 then UCL = 1; otherwise UCL.

The given mean value is used as a center line for all the chart points. However, the LCL and UCL are calculated for each chart point.

Sample Size: 15

CL = 0.6 (Mean/proportion) where 0.6 is the given mean value.

LCL = 0.6 – (3 * ) = 0.22053

UCL = 0.6 + (3 * ) = 0.97947

np

CL = Std. Mean (Given) * n

LCL = CL – (3 * )

UCL = CL + (3 * )

where n is the actual sample size from the sample, and Std. Mean (Given) is the given mean value.

If LCL < 0 then LCL = 0; otherwise LCL.

The CL, LCL and UCL are calculated for each chart point.

Sample Size: 15

CL = 0.6 * 15 = 9.0

CL = 9.0

LCL = 9.0 – (3 * ) = 3.30790

UCL = 9.0 + (3 * ) = 14.692099788303082

UCL = 14.69210

c

CL = Given Avg. No. of Defects per Unit * n

LCL = CL – (3 * )

UCL = CL + (3 * )

where n is the actual sample size from the sample, and GivenAvgNoOfDefects is the given average number of defects per unit (std_avg) for this characteristic.

If LCL < 0, then LCL = 0; otherwise LCL.

The CL, LCL and UCL are calculated for each chart point.

Sample Size: 15

CL = 10 * 15 = 150

CL = 150

LCL = 150 – (3 * ) = 113.25765

UCL = 150 + (3 * ) = 186.74235

u

CL = Given Avg. No. of Defects per Unit

LCL = CL – (3 * )

UCL = CL + (3 * )

where n is the actual sample size from the sample and GivenAvgNoOfDefects is the given average number of defects per unit (std_avg) for this characteristic.

If LCL < 0, then LCL = 0; otherwise LCL.

The LCL and UCL are calculated for each chart point.

Sample Size: 15

CL = 10.0

LCL = 10 – (3 * )

= 7.55051

UCL = 10 + (3 * )

= 12.44949

DPMO

where n is the actual sample size from the sample, AvgNoDPU is the given average number of defects per unit for this characteristic (std_avg), and NoDefectsOpps is the number of defects opportunities (num_defects_opp) configured for this characteristic.

If LCL < 0, then LCL = 0; otherwise LCL.

The LCL and UCL are calculated for each chart point.

Sample Size: 15

TitleResults for “How to create a CRG?”Also Available in