Six Sigma

SIX SIGMA

Six sigma is a quality standard, and not a technique. Techniques are adopted to reach that much quality. This is the benchmark, which every company is trying to achieve.

Sigma (s) is the standard deviation from the mean, or the required dimension in terms of manufacturing. It can also be implied to service sector, where mean will be related to quality of service or delivery time.

s can be either +ve or -ve. And variance= s2.

( Frnds, I cannot

draw........... so try to understand like this. )

think of the bell shaped curve. The center is the mean, where it has maximum amount of outputs. Any point in the lower area of that curve gives u the amount of job produced of a particular dimension. Take x-axis as different dimensions, n y-axis as number of jobs of that 'x' dimension. For example, imagine u hav to produce a stick of 5cm diameter. And ur tolerance is 0.5cm (hypothetical! So much variation on 5cm isn't acceptable, ever). Now the mean or middle of x-axis will represent 5cm diameter. Suppose max pieces produced are of 5cm, then at x=5 (or middle of x-axis) will hav max amount at y-axis. That's, the peak of bell-curved is highest at x=5. but, many pieces will be of either smaller or larger than 5cm diameter. Thise which are of more than 5cm dia are counted n marked on right side of x=5, n those which are less than 5cm dia are marked on left side. This way , u wil get a bell-curve, highest at the middle, n lower on sides................ Declining slope, first steep than lesser steep. As the accepted dimension as per specification is 5+/-0.5 cm, so those pieces which are of dia 4.5cm to 5.5cm are accepted, while others r rejected, as waste or scrap. Let the company producing it is working on 4 sigma, then the point on x-axis which is representing 5.5cm wil also be marked as +4s, n the point on x-axis which represent 4.5cm wil represent -4s. So the whole region is 8s, but it is called that the company has achieved 4s quality standard. Actually, s is calculated by formulae. The more the value of calculated s by formulae, the lesser amount of s-level is the quality standard of the company, because what is acceptable range of dimensions is more or less fixed. And the more is the standard deviation, the more is the wastage, as the bell curve will be more n more spread, n there will be more of pieces which will be beyond the acceptable lower (4.5) n higher (5.5) limits of x axis. The above two are called lower n upper specification limits, i.e., LSL (4.5) n USL (5.5).

now, divide the right hand side of the x-axis by 4, n u wil get the 1s,2s,3s,4s marks. Same with right side til -4s. As the LSL n USL are fixed, to reach higher slevels, means to actually lower the standard deviation frm the mean or the required dimension., so dat the USL can be made 5s and then 6s. So, minimize the s by different method improvements, more scientific n statistical.

As the s decreases, more n more jobs come under the range of LSL to USL, so the peak of bell increases n the spread of bell decreases. That means, more of pieces r are near to the mean dimension, n the number of rejections decreases as the bell has also shrunk.

Some facts: out of every 1,000,000 pieces produced

1 sigma 697,672 are rejected.

2 sigma 308,537 are rejected