Note - All products are only available to OEM source code purchasers. Contact us for details.

Use it for normal Statistical Process Control (SPC), or Six Sigma

Minimum System Requirements:

• 155 Mhz Pentium with 16 Meg. Ram
• 20 Meg. Hard Drive
• 640x480 Screen Size

The Features You Want

• Cut and paste charts and data into other software
• Uses a Microsoft Access97 Database
• Import/Export CVS file
• Very Easy to use

We have initiated new concepts into Six Sigma SPC software that no one else had. Mostly for Six Sigma reasons. This software is based on our Motorola SPC, (Statistical Process Control),  training that was the first training that we went to before the real six sigma training started. This is because everyone needed t know basic SPC to start. That training course was based on Dale H., Ph.D. Besterfield's book Quality Control the 2^nd edition and updated from the 6^th edition of Besterfield’s book, as well as Quality Planning and Analysis by Frank M. Gryna and Joseph M. Juran and it holds true to all the examples in the Motorola training materials. It deals with using process control charts.

• It is the easiest to use because there is no set up involved. It uses tab cards and Excel type grids that we are all familiar with. This means little to no time in training. So if you have a high turnover of employees or just move people around a lot, there is still no training involved.
• No classes to take to get your full value from it and you can remember how to run every aspect of it no matter how often you actually sit down at the keyboard. Plus for unskilled labor, 5 minutes or less of training and they can run every portion of it so you don't have to. It is cheap enough to allow everyone to use the same software thus eliminating confusion. This along can get you a return on t he investment in less than 3 months.
• Cp and Cpk, (process capability), line charts with not only the data plotted, but also your average Cp or Cpk, the Standard Deviation, (Std. Dev.), of the plots, your target design margin, three-sigma level, and a best-fit line. This can directly help you in tracking your Six Sigma projects.

With these charts, as well as the Range and X Bar charts, it can be difficult to actually see if you are improving or not. The ‘Best Fit’ line shows you this. You can estimate sigma all you want, but these charts are based on your actual data and actual calculations of sigma, not estimated. You can see flocculation in your process capabilities. Over time this can tell you how often you need to sample just for one think.

Let’s face it, if your X-Bar chart looks normal, it looks like a saw tooth, and you cannot really tell if it is going up or down. Our charts can tell you that.

Pre-Control Limits for X-bar and Range charts.

Data stored in Microsoft Access database for use with other products.

Export Data to a Microsoft Asscess97 Database or a .csv, (comma delimited) files.

Create as many databases as you need.

Save the Graphic Charts to the Hard Drive or Clipboard.

Correlated with a TI-36X Solar Statistics Calculator and Windows Calculator in Scientific Mode. TI 36X Solar Statistics Calculator rounds up to 5 at the tenth digit. We use double precision numbers.

Stand-A-Lone design. Does not require Microsoft Office or any Microsoft Office products, but will interface with them.

Easy data entry using the TAB or ENTER/RETURN key to enter you sample and move to the next sample.

Imports .csv, (comma delimited) files.

Our 3 tab card design makes it easy to browse and find a part number, parameter or sample data.

Windows 'What's This Help' for every field of every screen.

Extensive HELP file with help for every field on every screen.

Distribution/Histogram chart that compares your actual data with the part specification limits.

• (I.E. Nominal +/- 3 Sigma and Mean +/- 3 Sigma)

Purchase ZeroRejects for only $150.00(USD)

What we did that stands out the most is our distribution chart. We were the first to plot a bell shape curve around the specifications and the data. Why, well because six sigma is about process shift. Why no one else thought of it before us I don’t know. It just came to me one day. But our distribution chart shows much more than that.

The software we wrote for Motorola was the first software ever to have the lines to indicate 4.5 sigma. We expanded on that so that our software can be used for shops that are not six sigma. As you can see below, the scale of our distribution is laid out in units of sigma. This does several of things for you.

First it is easier to tell what you process is capable of producing compared to your specification limit. It tells you what you specification limit must be in order to achieve the specific design margin. But that is obvious. However, it does sort of give you a visual representation of the Cp and Cpk.

It also makes it quick to get a ‘Z’ number. Consider the graphic below.

This deals with the relationship between standard deviation and the area under the normal curve. You use a table found in most statistics books, but in order to use the table, any given ‘X’ value or measurement must first be converted to an appropriate ‘Z’ value. This is nothing more than the number of standard deviation units from the population mean.

Relating ‘Z’ values to ‘X’ Values

‘Z’ units are nothing more than numbers of standard deviation units. For example, one standard deviation to the right of x-bar is a value of 115 to t he right of x-bar. (sigma = 15) this also corresponds to a ‘Z’ value of one. A value of 122.5 would correspond to +1.5 ‘Z’ units. (1.5 standard deviations to the right or a value of 22.4 plus 100)

Of course there is a formula to do it also, but some times you just need a quick fix.

New in version 3 is a Cumulative Frequency chart. This is useful for checking if your distribution is normal or not. This is advantageous when doing a capability study.

According to my course example that was in Besterfield’s book verbatim (note the 6th edition of Besterfield’s book has another example that is similar) You can use an estimate of X-Bar and sigma, our software uses actual calculations instead.

First look at the distribution we invented, then the cumulative frequency chart.

NOTE: X-Bar above is 3.5

Estimate the average value by reading x value at the 50% point. This is the estimated mean and is about 3.0 compared to our calculated value of 3.5. This is actually the medium or center point.

Next consider the graphic below.

The lines on the chart are laid out like the lines on the image. So 68.26, total width between –1 and +1 sigma divided by 2 = 34.13. So we add and subtract 34.13 from the 50% middle line. And come up with 84.13 and 15.87. This is the spread of one sigma. Note that the standard deviation lines are not lined up at 84.13 and 15.87. This tells us something about our distribution. So does the fact that the Cf line intersects in more than one place the ‘Best Fit Line’.

Any percentage of values that could be expected below or above any given ‘X’ value can be read directly from the curve. We also do the 2 sigma spread the same way, but ones confidence at this extreme edge is less than the one sigma spread.

One advantage of ZeroRejects is that why you are looking at the charts on t he screen, the status bar at the bottom of the software always gives you the parameter specifications. This allows you to compare the numeric values on t he chart with the specification limits.

(Note: the status bar example does not match the data of the example charts. It is only shown for you to see how you can look at the charts on the screen and then look at the parameters all at the same time.)

The general interpretation of the probability plots are as follows.

If the distribution curve is normal, (or close to it), the resulting curve is a straight line, as shown in the example above. The average, spread and percentage beyond limits are easily obtained.

If the distribution is skewed, (lop-sided), the curve is not normal. However, the 50% point, (median), is valid; the percentages beyond specifications are valid; and the spread is valid.

By looking at where the line starts and finishes vs. the specification and how straight it is compared to the best fit line, and where the standard deviation lines are, you can picture if your distribution is normal or not graphically.

User selectable Design Margin (Target Sigma) 3 to 6 Sigma.

Smart Evaluation of Design Margin VS Actual Data so every station doesn't need 'Six Sigma' training

X Bar, Range, Cp and Cpk Line Charts for 6 Sigma, ISO TQM, and QS requirements. All line charts have a 'Best Fit' line so you can see the trends.

X Bar

Range

Cp

Cpk

Last Updated: Sunday, 13-Apr-08 07:22:06 PDT