Control: SPC, Control Plan, and Handover
Control: SPC, Control Plan, and Handover
Control sustains the gain and transfers ownership. The exam tests control-chart selection, control limits versus specification limits, out-of-control rules, reaction plans, standard work, and project closure.
Core rules and concepts
- Control plan: critical Xs and Ys, specification, measurement method, sample size and frequency, chart type, reaction plan, and owner.
- Chart selection: Xbar-R for subgrouped continuous data; I-MR for individual data; p, np, c, or u for attribute data depending on defectives versus defects and constant versus variable sample size.
- Control limits come from process variation, not customer specifications. Plot specifications only as reference lines.
- Out-of-control signals: a point beyond a control limit, a run of seven on one side, a trend, or other Western Electric rules. Investigate special cause; do not recalculate limits to hide it.
- Reaction plan (OCAP): contain suspect product, identify cause, correct, verify, and record.
- Sustain: standard work, visual management, training, layered audits, poka-yoke, and a named process owner.
Worked example
An Xbar-R chart on fill volume shows a subgroup mean above the upper control limit. The reaction plan says stop, inspect the fill nozzle, check calibration, and segregate product since the last good check. Reasoning: the signal is special cause; recalculating limits would conceal it. After correction, confirm stability and capability, then update the control plan.
Traps that cost marks
Putting specification limits on a control chart; recalculating limits after every point; treating common-cause variation as special cause; having no reaction plan; closing the project before the process owner accepts it; reporting savings without verification; using an Xbar-R chart for individual data.