"It's Too Complex for me to Understand!!"............. Quote from I.M.Awiner, SMS Manager..L.A.Y.Z. airways.
Every Safety Management System requires the company to “monitor” processes to assure that they are in control. Most companies accomplish this through the Audit process required by a Quality Assurance System. The question arises:
Once a month, once every 6 months, or once a year. What about the times “in-between” the audits? This is where Statistical Process Control, SPC, comes in. The entire SPC system is based on a “normal” distribution of data. We have discussed what is “normal” in other articles in this blog. Through the use of statistics, we have the ability to “Quantify” the “Variation” from what is considered “Average" in any process. This variation is encompassed in statistical boundaries called “Control Limits.” These control limits are measured in Standard Deviations which are commonly refered to as “SIGMAs.”
|This is an illustration of the Bell Shape Normal Distribution Curve|
with Superimposed Standard Deviations or Sigmas.
For those who are mathematically inclined, the Standard Deviation, or Sigma, is a distance that can be calculated. Using the average or mean of the data, we can simply plug that information into a formula that calculates the Standard Deviation or Sigma. In the case of percent defective, the Standard Deviation or Sigma is represented by the formula:
|The Formulas for Calculating the Upper and Lower Control Limits|
to a Percent Defective Control Chart. Formulas found in the SMS Memory Jogger II.
Let's take a look at the Process of measuring the output of a process usisng Statistical Process Control and the Control Chart tool.
All of the Metrics, or Data, must fall within an area, 3 times one sigma up from the mean and 3 times one sigma down from the mean giving us an “In-Control” area that is 6 sigmas big. Thus the term 6 sigma Quality Control. It is important to remember that these control limits simply tell us if the process is “Normal" or "In Control." The Control Chart does NOT tell us if we are producing a good output.
By simply recording metrics from any process, then putting that data into a distribution and then calculate control limits. We can tell is the process is normal or not. If the process is normal and none of the data points stay within the control limits, we then have the power to predict what that process will do in the near future....we can plan!
|This is the Six Sigma Symbol commonly used.|
Think about the Control your company would have if we could monitor processes using the 6 Sigma method. Your probably saying now, holy crap, I can’t do math, I will never be able to do this. Rest assured that most companies do not have statistician that could implement this system. There are simple computer programs that will do this for you and these programs are surprisingly affordable. I like “SPC for Excel” which available on the web. For under $100, you can have the control power of 6 sigma. I am sure if you Goggle other SPC program, you will find others that are affordable.
As far as SMS and 6 sigma....If your company is using control charts properly, your Safety Management Quality Assurance System will be acting on variation rather than incidents or accidents. A truly pro-active system.........your thoughts.
(NOTE: Additional information on SPC and Control Charts is found in the SMS Memory Jogger II authored by myself and Sol. You can get your copy at dtitraining.com )