Defect rates look small until they are scaled. A process with one defect in a hundred chances sounds fine, yet it produces ten thousand defects per million. This DPMO calculator turns defects, units inspected and opportunities per unit into DPMO, DPU, yield and the sigma level.
How the calculation works
- DPU = total defects ÷ units inspected.
- DPO = total defects ÷ (units inspected × opportunities per unit).
- DPMO = DPO × 1,000,000.
- Yield = (1 − DPO) × 100, the share of opportunities without a defect.
- Sigma level = the normal quantile of (1 − DPO) plus 1.5 for the usual long-term shift. At six sigma the DPMO is 3.4.
Example
An inspection of 500 units, each with 5 chances for a defect, finds 27 defects. DPU = 27 ÷ 500 = 0.054, DPO = 27 ÷ (500 × 5) = 0.0108 and DPMO = 10,800. The yield is 98.92 % and the sigma level is about 3.80, a process that is better than average but far from six sigma.
Why it matters
Quality teams in manufacturing, software, healthcare and services use DPMO to compare processes of different size and complexity on one scale. The sigma level translates a defect rate into a number that managers know, and a project can show progress by moving from 3.5 to 4 sigma in a measurable way.
Good to know
DPMO depends on how opportunities are defined: counting more opportunities per unit lowers the DPMO without improving anything, so the definition must be fixed and honest. The 1.5 sigma shift is a convention with debated support, so state whether you use it. Small samples give wide uncertainty, so a zero-defect sample does not prove a perfect process. Use DPMO together with process capability and control charts. Teams often track it for a few critical steps rather than the whole product, which keeps the numbers meaningful and the improvement work focused.
Related tools
The process capability calculator measures Cp and Cpk, the control chart calculator sets limits for monitoring a process and the confusion matrix calculator scores classification errors.