Out-of-control action plan (OCAP)

Why an OCAP

A control chart only helps if someone reacts to it, and reacts the same way every time. Without a plan, one shift adjusts the machine at every signal, another ignores signals, and a third stops and waits for an engineer. An out-of-control action plan (OCAP) is the agreed reaction, written as a flowchart and posted with the chart. It is the reaction plan of the control plan, made specific.

The three parts

• Activators: the signals that start the plan. These are the chart rules in use: a point beyond a control limit, a run on one side of the center line, a trend, and so on. Agree which rules apply to this chart; using every rule on every chart raises false alarms.
• Checkpoints: yes/no questions about the likely causes, in order. Put the quickest and most likely first. The first is almost always the measurement: was the part measured, recorded and plotted correctly?
• Terminators: the actions each finding leads to, ending in "resume production" or an escalation to someone who can investigate further.

Every OCAP also needs to say what happens to the product made since the last good point. A process that was out of control may have made bad parts before anyone noticed.

Writing the steps

Write each check so that Yes means "found it" and leads to an action, and No moves to the next check. Name who does each step and how long it should take. Keep the chain short: a long list is skipped in practice. If the checks find nothing, escalate; do not let the operator guess.

The tool checks that every signal leads somewhere, every check has both answers, every ID exists, every step can be reached, and every path ends at "resume" or an escalation, with no loops that never end.

The event log

Each use of the plan is one line in the log: the signal, where the cause was found, what was done and how many parts were suspect. The log turns reactions into learning. When the same cause shows up again and again (an insert that chips every two weeks), the OCAP is doing its job, but the process needs a corrective action that stops the cause, such as a tool-life limit. When many events end in escalation with no cause found, the plan is missing a check.

On the exam

CQPA III.E.5 and CQE VI.F.6 cover reading control chart patterns and the rules for statistical control. CSSBB VIII.A.5 covers interpreting control charts to tell special from common causes, and VIII.C.2 the control plan that keeps the improved process in control. CMDA V.A includes setting alert and action levels and SPC charts. Expect to choose the right reaction to a given signal, and to recognize that adjusting a process in response to common-cause variation (tampering) makes it worse.

Common questions

What is an OCAP?

An out-of-control action plan is the predefined response to a control chart signal: what the operator checks first, what to do for each finding and when to stop and escalate. It turns "investigate the cause" into steps anyone on the shift can follow the same way.

What goes in an OCAP flowchart?

Three kinds of box: the activator (the signal, such as a point beyond a control limit or a run), the checkpoints in the order they should be checked (measurement, material, setup, tooling), and the terminators (the actions: adjust, contain, stop the line, call engineering). This page builds the flowchart as you write the steps.

How is an OCAP different from the reaction plan in a control plan?

The Control plan column says in a line what happens when a characteristic goes out of control or out of specification. The OCAP is the detailed decision flow behind that line.

What counts as an out-of-control signal?

A point beyond a control limit is the main one. Most plants also act on nonrandom patterns such as a run of points on one side of the center line or a steady trend, using the run rules they have adopted. Write the rules you use into the activator so everyone reacts to the same signals.

Other tools and templates