DMS Statistics

This setting affects how GainSeeker calculates the statistics Total Sample Size, Total DPM, Total DPB, Total DPK, % Defects, % Good (not defects), and Total good Units. In the Legacy modules Enterprise Dashboard and DMS Charts and Reports, it also affects the statistics Yield, % NCU, and % Good Units.

When this check box is cleared (the most typical setting), these statistics are calculated relative to the total units inspected (sum of the sample sizes from all the data records in the current retrieval).
When this check box is selected, GainSeeker uses the Opportunities per Unit from each DMS standard, multiplied by the total sample size retrieved for each standard, to calculate these statistics relative to the total opportunities for error instead of the total units inspected. This is useful in situations where different parts have vastly different levels of complexity.

Data type

You can select to force the system to always regard data as a certain type (p, np, c, or u), or you can have GainSeeker automatically determine the data type.

DPM calculations

Configure options for calculating DPM (defects per million).

Method to calculate DPM

Modify the way defects per million is calculated to better estimate when you have small sample sizes.

For information about the Standard method, see Calculating DPU, DPMO, PPM and related values
For information about the Best estimate method, see Alternate methods of estimating DPM

Method to handle DPM across processes

When you retrieve defect data across processes, you can calculate defects per million using two different methods: the Averaged Method or the Additive Method. These different methods can produce very different results in certain situations. The averaged method is preferred by most customers, especially during start up, because it is less sensitive to new processes being added to the retrieval.

Never report zero

See Alternate methods of estimating DPM.

User conservative calculation

See Alternate methods of estimating DPM.

DPU improvement goal

You can set a customized improvement goal for the DPU 10x charts. The default is 10x (times) improvement over two years. You can make that more aggressive or more conservative depending on your situation.

Exclude statistical outliers

You can select to include or exclude statistical outliers from all calculations.

First of five traceability columns where OEE values are stored =

Enables OEE calculations when the traceability field label Available time is set. Not set is the default option.

For more information about this and other simple prerequisite tasks you must perform before retrieving OEE data, see OEE Retrieval prerequisites

Number of decimals for a cost value

Change the number of decimals you want to display for cost values.

Number of decimals for a number

Change the number of decimals you want to display for number values.

Use cost information from

Choose how GainSeeker uses cost information.

Bypass cost information by selecting the Bypass cost to improve speed option.
The Defect list option uses individual defects costs.
The Standard list uses NCU (nonconforming unit) costs.
If you choose to have GainSeeker use cost information from a Traceability column, use the Use traceability column for cost information setting below to select which traceability to use. This is especially useful for times when GainSeeker integrates with other systems.

Use traceability column for cost information =

Select which traceability column to use when you select the Traceability column option for the Use cost information from setting.

See important Notes about using this setting.

Notes about the Use traceability column for cost information setting

To ensure proper calculation of the statistics:

Select the checkbox for the Set nonconforming units = number of defects setting. (Additionally, if this checkbox is not selected and specific defects are selected during the retrieval, nonconforming unit costs other than zero may be returned even if the number of defects is zero.)

Storing more than one defect reason in each data record and selecting specific defects during the retrieval:

Returns the total cost stored with the data. (A percentage based on the number of defects and the total defects is not calculated.)

Selecting multiple defect reasons to be entered in one data record:

Splits the cost proportionately among defects. Example: If the cost is $10 and there are five defects (three broken and two blistered), the cost of each defect is assumed to be $2. The cost is not divided unevenly for defects with varying costs in one data record. For example, the cost of three defects (broken equals $3) and two defects (blistered equals $7). You can accomplish this by splitting the data into two data records.

Also see Cost from statistics and Assigning defects