I85PREP
⌘ K
CMRP study materials
CMRP study resources

CMRP: Manufacturing Process Reliability

CMRP · Work through your exam, one topic at a time.

All guide sections

Manufacturing Process Reliability

What this domain tests

Manufacturing process reliability examines the stability and capability of the conversion process, not only the equipment. Questions cover OEE, process capability, statistical process control, bottleneck and constraint management, changeover, and the interaction of people, method, material, machine, and measurement.

Core rules and concepts

  • OEE = Availability × Performance × Quality. It uses planned production time, so planned downtime is excluded but unplanned stops and speed loss are not.
  • Separate equipment losses from process losses. A reliable machine feeding a poorly controlled process still produces scrap.
  • Capability indices: Cp compares specification width with process spread; Cpk also accounts for centering. A process can be capable in spread but off-center.
  • Use control charts to distinguish common-cause variation from special-cause variation. Reacting to common-cause noise creates tampering.
  • Manage the constraint. Improving a non-bottleneck step does not raise throughput unless it relieves the constraint.
  • Changeover time is a process reliability issue: long, variable changeovers create startup defects and lost capacity.

Worked example

Planned production time is 480 minutes; unplanned downtime is 48 minutes; ideal rate is 1 unit per minute; total output is 400 units; good output is 380. Availability = (480 − 48) ÷ 480 = 90.0%. Performance = 400 ÷ (432 × 1) = 92.6%. Quality = 380 ÷ 400 = 95.0%. OEE = 0.90 × 0.926 × 0.95 = 79.2%. Reasoning: the 48 minutes is excluded from the performance base only through the availability calculation; performance compares actual output with what the remaining 432 minutes could produce at ideal rate. The largest recoverable loss here is quality at 5%, followed by speed loss at 7.4%.

Traps that cost marks

Using total calendar time instead of planned production time. Counting rework as good output. Treating OEE as a pure maintenance metric. Assuming a high Cp guarantees a high Cpk. Improving a non-constraint and expecting throughput gains. Using specification limits as control limits on a control chart.

Exam tip

When OEE is low, decompose it before choosing an action. A 60% OEE driven by availability requires different work than 60% driven by quality.