CMRP cheat sheet

Return on Investment (ROI) = (Gain from Investment − Cost of Investment) / Cost of Investment; express as a percentage to compare competing reliability projects.

Overall Equipment Effectiveness (OEE) = Availability × Performance × Quality; world-class benchmark is commonly cited as 85% (90% × 95% × 99.9%).

Availability = Operating Time / Planned Production Time = MTBF / (MTBF + MTTR); raising MTBF and cutting MTTR both improve availability.

Maintenance cost per unit produced = total maintenance cost / units produced; use it to track whether reliability spending scales with output.

Life-Cycle Costing sums acquisition, operating, maintenance, and disposal costs; a cheaper purchase price can lose to a higher-reliability asset over its life.

Budget categories: reactive/corrective, preventive, predictive, and improvement work; a mature program shifts spend from reactive toward planned work.

Payback period = initial investment / annual net savings; shorter payback reduces financial risk but ignores cash flows after the payback point.

Key performance indicators should be paired (leading with lagging); cost per unit and OEE are lagging, PM compliance and schedule compliance are leading.

Process capability Cp = (USL − LSL) / 6σ measures spread only; Cpk = min[(USL − μ), (μ − LSL)] / 3σ also accounts for centering.

Cpk of 1.33 corresponds to roughly 63 defects per million opportunities; 1.67 corresponds to roughly 0.6 defects per million.

A process must be statistically in control before capability indices are meaningful; special-cause variation invalidates Cp/Cpk.

Critical process parameters are the inputs whose variation drives product variation; control them rather than inspecting output only.

Poka-yoke (error-proofing) prevents defects at the source by making the wrong action physically impossible or immediately obvious.

Six Sigma DMAIC sequence: Define, Measure, Analyze, Improve, Control; Control is where gains are standardized and handed to operations.

Total Productive Maintenance targets zero breakdowns, zero defects, and zero accidents by building operator ownership of equipment condition.

Process FMEA scores Severity × Occurrence × Detection; higher Risk Priority Numbers drive priority for corrective action.

MTBF = total operating time / number of failures; MTTR = total repair time / number of repairs; both require a defined failure and time boundary.

Failure rate λ = 1/MTBF for the constant-rate (useful life) region; reliability R(t) = e^(−λt) for that exponential model.

The bathtub curve has three regions: decreasing early failures (infant mortality), constant random failures, and increasing wear-out failures.

Weibull shape parameter β < 1 indicates infant mortality, β ≈ 1 random failures, β > 1 wear-out; scale parameter η is the characteristic life at 63.2% failure.

Reliability block diagrams: series reliability multiplies component reliabilities; parallel (redundant) paths raise system reliability.

Age-based replacement is justified only when β > 1 (wear-out); for random failures, condition-based or opportunistic replacement is preferred.

Root Cause Analysis tools: 5 Whys, fishbone (Ishikawa), and fault tree analysis; stop at the physical, human, and latent (system) causes.

RCM decision logic asks: what are the functions and performance standards, what causes functional failure, and is a proactive task technically feasible and worth doing?

Span of control is the number of direct reports per supervisor; wide spans flatten cost but reduce coaching depth.

Reliability culture is measured by behaviors and metrics, not slogans; leaders set priorities when schedule and maintenance conflict.

Change management follows unfreeze–change–refreeze; communicate the why, build a guiding coalition, and lock in gains.

Cross-functional reliability teams typically include operations, maintenance, engineering, planning, and stores; shared metrics prevent silo optimization.

Accountability requires a single owner per asset or system; shared ownership without a named owner reliably produces gaps.

Succession and skills-gap analysis compare required competencies to current capability to target training investment.

A steering committee governs priority and funding for improvement projects; without governance, improvement work loses to daily firefighting.

Work order lifecycle: request, screening/prioritization, planning, scheduling, execution, closeout with failure coding, and analysis.

Planning defines scope, labor, materials, tools, permits, and duration before scheduling; scheduling assigns work to a time slot and crew.

Weekly schedule compliance = jobs completed as scheduled / jobs scheduled; a common target is 90% or higher.

Backlog is measured in crew-weeks of ready work; a common planning guideline is 2–4 weeks of ready backlog to absorb variation.

Priority codes should reflect consequence (safety, environmental, production, cost), not the loudest requester.

Planned and scheduled work typically consumes less labor per job than reactive work because travel, waiting, and parts delays are removed.

Job closeout must capture actual labor hours, parts used, and a failure code; without it, history and future planning degrade.

Emergency work should be a small share of total hours; a rising emergency ratio signals weak planning or deferred PM.

Preventive maintenance is time- or usage-based (calendar intervals, run hours, cycles); predictive is condition-based and triggers work when a measured parameter crosses a limit.

Vibration analysis detects imbalance, misalignment, looseness, and bearing defects; high-frequency techniques catch early bearing faults.

Thermography finds electrical hot spots, insulation degradation, and friction heating; it requires a load and a reference baseline.

Oil analysis tracks wear metals, contamination, viscosity, and water; trending over samples matters more than a single result.

Ultrasonic testing detects compressed-gas leaks, steam trap failures, and early mechanical friction at high frequency.

PM optimization removes tasks that do not address a credible failure mode; over-maintenance introduces infant mortality and consumes capacity.

Condition monitoring interval should be short enough to catch the failure before functional failure; set it from P-F interval data.

Potential failure (P) to functional failure (F) interval defines the window for predictive detection; the inspection interval should be a fraction of it.

Economic Order Quantity EOQ = √(2DS/H), where D is annual demand, S is order cost, and H is annual holding cost per unit.

Safety stock buffers demand and lead-time variability; higher service level targets require larger safety stock.

Reorder point = (average daily demand × lead time in days) + safety stock.

ABC classification ranks inventory by annual usage value; A items get tight control, C items get simple periodic review.

Critical spares are identified by consequence of unavailability and lead time, not by unit price alone.

Storeroom accuracy is verified by cycle counting rather than annual wall-to-wall counts; count A items most frequently.

Vendor-managed or consignment inventory shifts holding cost to the supplier but requires clear consumption reporting and reconciliation.

Obsolescence review removes stock for retired assets; unused inventory ties up cash and hides true demand signals.

Training needs analysis identifies the gap between required and current competency at organization, job, and individual levels.

Competency models define observable skills and knowledge per role; certification should require demonstrated performance, not attendance alone.

On-the-job training with a qualified mentor transfers tacit skills; pair it with a structured checklist and sign-off.

Kirkpatrick evaluation levels: reaction, learning, behavior, and results; only the last two show on-the-job impact.

Cross-training reduces single-point-of-knowledge risk; identify critical skills held by only one person as a priority gap.

Training records support audit and regulatory compliance; track completion, renewal dates, and qualification expiry.

Procedures should be written at the reading level of the audience and validated by a trial run by a qualified non-author.

Hierarchy of controls, most to least effective: elimination, substitution, engineering controls, administrative controls, then personal protective equipment.

Lockout/tagout requires isolating all energy sources, verifying zero energy, and applying personal locks; one lock per worker.

Job hazard analysis breaks a task into steps and identifies hazards and controls before work begins.

Permit-to-work systems control hot work, confined space entry, and line breaking; permits must be verified on site, not just signed.

Incident metrics: total recordable incident rate = (recordable cases × 200,000) / total hours worked.

Near-miss reporting is a leading indicator; low reporting often signals fear rather than safety.

Spill prevention includes secondary containment, drip pans, and correct labeling; environmental reporting may be legally required.

Security controls include access control, visitor logging, and control of hazardous or high-value materials in the storeroom.

PDCA cycle: Plan, Do, Check, Act; standardize the improvement in Act so the gain is sustained.

Lean waste categories: overproduction, waiting, transport, overprocessing, inventory, motion, and defects (plus unused talent).

5S sequence: Sort, Set in order, Shine, Standardize, Sustain; it is a foundation for visual management.

Kaizen is small, frequent, team-based improvement; kaizen events compress change into a short focused window.

Value stream mapping identifies value-added versus non-value-added steps and quantifies lead time versus process time.

Standard work defines the current best method with cycle time, work sequence, and required inventory; it is the baseline for further improvement.

A3 problem solving captures background, current state, target, root cause, countermeasures, and follow-up on one page.

Improvement projects should be tied to a measurable baseline and verified result; without a baseline, savings cannot be validated.