kontrollimit

Kontrollimit, or control limit, is a threshold used in statistical process control to delineate the boundary of expected variation in a process that is considered to be in statistical control. It represents the natural variability of the process, not customer requirements. The concept is used on control charts to detect when the process may have shifted due to assignable causes.

Control limits are computed from historical process data. In many practical charts, upper control limit (UCL) and lower control limit (LCL) are set at the mean plus or minus three standard deviations, often referred to as the three-sigma rule. For example, if a process has a mean diameter of 10.0 mm and a standard deviation of 0.05 mm, the X-bar control chart would have UCL = 10.15 and LCL = 9.85.

Control limits differ from specification limits, which are derived from product requirements. A point outside control limits suggests an assignable cause and the process may be out of statistical control, while a point within limits does not guarantee conformity, only that the observed variation is within expected boundaries given the model.

Beyond X-bar charts, other chart types use their own limit calculations, such as p-charts for fractions defective and c-charts for defect counts; in all cases the limits are designed to reflect the expected variation given the sample size and distribution assumptions.

Interpreting a chart involves looking for points outside the limits, runs, trends, or unusual patterns that indicate special causes. If such signals appear, investigators seek root causes and take corrective actions, then recalculate limits after restoring process stability.

Historically, control charts were introduced by Walter A. Shewhart in the 1920s and form a core tool of quality assurance in manufacturing and service processes.