Total productive maintenance

In short

Total productive maintenance (TPM) is a company-wide system, developed by Seiichi Nakajima and promoted by the Japan Institute of Plant Maintenance (JIPM), that aims at zero breakdowns, zero defects and zero accidents by making every employee — especially machine operators — responsible for the health of equipment. It matters because in capital-intensive Indian plants the hidden losses of idle, slow and defect-producing machines usually cost more than labour, and OEE is the single number managers use to expose them. It is examined in the Maintenance Engineering / Industrial Engineering / Quality and Reliability papers of most AICTE-pattern B.E. and B.Tech programmes and in the GATE PI and ME syllabi (maintenance, reliability, availability, OEE). Indian manufacturers such as Tata Steel, Bajaj Auto, Hero MotoCorp, Mahindra, Godrej, Sundaram Clayton, Rane and TVS (several JIPM TPM Excellence Award winners) run formal TPM programmes, so OEE and pillar questions are standard in plant and maintenance interviews.

Explained

Maintenance strategies and where TPM sits

• Breakdown (reactive) maintenance: repair after failure. Cheapest to plan, costliest in downtime, overtime and consequential damage; acceptable only for cheap, non-critical, redundant items (a spare bulb, one of many fans).
• Preventive maintenance (PM): time- or usage-based servicing (every 500 hours, every 10,000 km). Simple to schedule, but replaces parts with life left and cannot catch random failures.
• Predictive / condition-based maintenance (CBM): act on measured condition — vibration spectra, oil (tribology) analysis, infrared thermography, ultrasonic leak detection, motor current signature analysis. Highest diagnostic value; needs instruments and skilled analysts.
• Corrective maintenance: repair plus modification so the same failure cannot recur (re-design, better material, guard, interlock).
• Maintenance prevention / design-out: build reliability and maintainability into new equipment — the Development Management pillar.
• TPM is not a fifth technique: it is the management framework that combines all of the above with operator ownership, small-group kaizen and measurement by OEE. RCM (reliability-centred maintenance), by contrast, is an analytical method that uses FMEA and failure consequences to *select* the right task for each function; TPM is people- and loss-driven, RCM is function- and risk-driven, and large Indian plants use both.
• Autonomous maintenance distinction: operators do cleaning, lubrication, bolt tightening, inspection and simple adjustments ("CLIT"/CLAIR tasks); the maintenance department keeps overhauls, precision work and diagnostics.

Definition, goals and the 5S / JH foundation

• Definition (JIPM): TPM is productive maintenance carried out by all employees through small-group activities, aiming to maximise equipment effectiveness over its whole life.
• The targets: zero unplanned breakdowns, zero defects, zero accidents, zero (minimum) inventory of excess spares.
• Three key ideas: total *effectiveness* (profit-oriented), total *maintenance system* (prevention + prediction + maintainability improvement), total *participation* (operators to top management).
• 5S foundation: Seiri (sort), Seiton (set in order), Seiso (shine — in TPM, cleaning *is* inspection), Seiketsu (standardise), Shitsuke (sustain). Visual controls — oil-level marks, match-marks on bolts, direction arrows, green/red gauge bands, one-point lessons — make an abnormality visible to a passing operator.
• Fuguai (abnormality) tags: red tags for maintenance-owned defects, white tags for operator-owned; tag counts closed per month is a standard TPM metric.
• Natural vs accelerated deterioration: wear is unavoidable; dirt, looseness, poor lubrication and wrong operation *accelerate* it. TPM attacks accelerated deterioration first because it is free to remove.

The eight pillars

1. Jishu Hozen (autonomous maintenance): operators own their machines through 7 steps — (i) initial cleaning, (ii) counter-measures to sources of dirt and hard-to-reach areas, (iii) tentative cleaning/lubrication standards, (iv) general inspection training, (v) autonomous inspection, (vi) standardisation and workplace management, (vii) full autonomous management.
2. Kobetsu Kaizen (focused improvement): cross-functional teams eliminate the 16 major losses, step by step, with PM analysis, why-why (5-why) analysis and the 7 QC tools.
3. Planned maintenance: build a machine register and criticality (A/B/C) classification, set PM frequencies, spares and checklists, and drive MTBF up and MTTR down; targets zero breakdowns.
4. Quality maintenance (Hinshitsu Hozen): set and hold the equipment conditions that cannot produce a defect — condition control, not inspection; uses the Q-M matrix, 4M (man, machine, material, method) analysis and poka-yoke.
5. Education and training: raise skill through a four-level matrix (knows nothing → knows theory → can do → can teach), one-point lessons and skill maps.
6. Office TPM (administrative): removes losses in purchase, stores, planning and accounts — order lead time, document retrieval, spares procurement.
7. Development management (early equipment / early product management): feed shop-floor lessons into new machine specification and layout so the next line starts up vertically (fast ramp-up, MP design).
8. Safety, health and environment: zero accidents and zero pollution; Hiyari-hatto (near-miss) reporting, Kiken Yochi Katsudo (KYK/hazard prediction), poka-yoke guarding. In India this pillar now maps to the duties of the Occupational Safety, Health and Working Conditions Code, 2020 (in force 21 November 2025, replacing the Factories Act, 1948) and to standards such as ISO 45001:2018 and IS/ISO 14001.

The six big losses and OEE

• OEE splits equipment time into three factors, each hit by two losses:
• Availability losses — (1) breakdown/failure (sudden and chronic stoppages needing maintenance) and (2) setup and adjustment (changeover, die change, warm-up, trial pieces). Attacked by planned maintenance and SMED.
• Performance losses — (3) idling and minor stoppages (jams, sensor trips under 5-10 minutes, cleared by the operator without any maintenance intervention) and (4) reduced speed (running below design cycle time to avoid defects or noise).
• Quality losses — (5) defects and rework in steady state and (6) start-up / yield losses at the beginning of a run or after a changeover.
• JIPM's wider list adds planned downtime, management, motion, line organisation, logistics, measurement and adjustment losses plus man, material, energy, die/jig/tool and yield losses — sixteen in all.
• Loading time excludes planned shutdowns (holidays, planned PM, no-order time); operating time = loading time − downtime. Two plants can quote very different OEE for the same machine purely by how much they put in "planned".
• TEEP (total effective equipment performance) measures OEE against all 24 × 365 calendar hours and exposes the capacity hidden in planned downtime.
• World-class benchmarks (JIPM, indicative): availability ≥ 90%, performance ≥ 95%, quality ≥ 99% → OEE ≈ 85%. Typical Indian job-shops start at 35-55%.

Reliability, maintainability and availability measures

• MTBF (mean time between failures) measures reliability; MTTR (mean time to repair) measures maintainability; MWT (mean waiting time for spares, permits, manpower) is often larger than MTTR and must be tracked separately.
• Inherent availability uses MTBF and MTTR only; operational availability also includes logistic delay and planned maintenance — exam questions usually mean the first unless delays are given.
• Failure rate λ = 1/MTBF for a constant-failure-rate (exponential) item; the bathtub curve has an infant-mortality (burn-in) region, a useful-life region of constant λ, and a wear-out region where time-based replacement pays.
• Maintenance cost indices: maintenance cost as % of replacement asset value (RAV) — good plants hold about 2-3% per year; PM-to-breakdown work-order ratio; planned maintenance percentage (target > 80-85%); schedule compliance.

Implementation roadmap, benefits and limitations

• JIPM's 12 steps in four phases: *Preparation* — (1) top management announces TPM, (2) launch education and publicity, (3) create TPM promotion office and pillar committees, (4) set basic policy and goals, (5) draw the master plan; *Introduction* — (6) kick-off involving customers and suppliers; *Implementation* — (7) raise equipment effectiveness (model machine first), (8) autonomous maintenance programme, (9) planned maintenance by the maintenance department, (10) skill training, (11) early equipment management; *Stabilisation* — (12) perfect TPM and apply for the JIPM award (Excellence Award → Consistency → Special → Advanced Special).
• Model machine approach: prove gains on one bottleneck machine, publish before/after boards, then roll out — avoids the "TPM everywhere, nowhere" failure.
• Advantages: higher OEE without capital spending, lower maintenance and spares cost, fewer defects, better safety, higher operator skill and morale, and the stable equipment that JIT/lean flow requires.
• Limitations: 2-3 years to mature; needs sustained management commitment and training time (production must release operators); union or supervisory resistance to operators doing "maintenance work"; heavy documentation; gains plateau unless kaizen continues; OEE can be gamed.
• Indian standards and practice: IS 15658 (maintenance management guidance) and the IS/ISO 55001 asset-management series are the usual references for maintenance management systems; ISO 9001:2015 clause 7.1.3 (infrastructure) is where auditors look for maintenance evidence (check the clause in ISO 9001:2026, published 16 September 2026, which succeeds the 2015 edition); IS/ISO 20816 series for machine vibration evaluation and IS/ISO 18436 for condition-monitoring personnel. Statutory periodic examination of pressure vessels, lifts and hoists, and pollution-control equipment runs under the Occupational Safety, Health and Working Conditions Code, 2020 and state factory rules, and these inspections cannot be traded away for OEE.

Key points

• TPM - plant-wide system for zero breakdowns, defects and accidents through total employee participation; originated at Nippondenso, codified by JIPM.
• Jishu Hozen - operator-owned cleaning, lubrication, tightening and inspection in 7 steps; cleaning is inspection.
• Eight pillars - JH, focused improvement, planned maintenance, quality maintenance, training, office TPM, development management, safety-health-environment, on a 5S foundation.
• Six big losses - breakdown, setup/adjustment, idling/minor stoppage, reduced speed, defects/rework, start-up loss.
• OEE - availability × performance × quality; the single number that converts the six losses into lost capacity.
• Availability - operating time ÷ loading time; planned downtime is excluded from loading time.
• Performance rate - (ideal cycle time × total pieces) ÷ operating time; catches speed and minor-stoppage loss.
• Quality rate - good pieces ÷ total pieces produced.
• TEEP - OEE × utilisation of calendar time; shows capacity hidden inside planned downtime.
• World-class OEE - about 85% (90 × 95 × 99); most Indian SMEs begin in the 35-55% band.
• MTBF and MTTR - reliability and maintainability measures; A = MTBF/(MTBF+MTTR).
• Accelerated deterioration - wear caused by dirt, looseness and poor lubrication; removing it is free and is attacked first.
• Fuguai tags - red/white abnormality tags; tags raised and closed is a live TPM score.
• Chronic vs sporadic loss - sporadic failures are sudden and obvious; chronic small losses are larger in total and need PM analysis.
• Quality maintenance - hold the equipment conditions that cannot make a defect, instead of inspecting defects out.
• TPM vs RCM - TPM is people- and loss-driven and bottom-up; RCM is FMEA- and consequence-driven and analytical.
• 12-step JIPM roadmap - preparation, introduction, implementation, stabilisation; typically 2-3 years to award level.
• MP design / early equipment management - designing maintainability and reliability into the next machine, the cheapest fix of all.

Formulas and examples

Example 1 — OEE of a CNC turning centre (one shift) Shift = 480 min; planned lunch and PM = 60 min; breakdown = 40 min; setup = 20 min; ideal cycle time = 0.5 min/piece; pieces produced = 600; rejects = 18.

• Loading time = 480 − 60 = 420 min.
• Unplanned downtime = 40 + 20 = 60 min → Operating time = 420 − 60 = 360 min.
• A = 360/420 = 0.857 (85.7%).
• P = (0.5 × 600)/360 = 300/360 = 0.833 (83.3%).
• Q = (600 − 18)/600 = 582/600 = 0.97 (97%).
• OEE = 0.857 × 0.833 × 0.97 = 0.6926 → 69.3%.
• Valuable operating time = 0.6926 × 420 ≈ 291 min; i.e. 129 of 420 available minutes are lost.
• Biggest single lever: performance loss (minor stoppages/speed) — 60 min equivalent, more than the 40-min breakdown.

Example 2 — where the losses sit, in minutes and rupees Same machine as Example 1, contribution lost = ₹25 per piece of lost output; ideal output in 420 min = 420/0.5 = 840 pieces; good pieces = 582 → shortfall 258 pieces.

Check: 582 good pieces × 0.5 = 291 min = valuable operating time, matching Example 1. Per 25-day month and 2 shifts, the same loss pattern is about ₹3.2 lakh.

Example 3 — MTBF, MTTR and availability before and after TPM A press runs 7,200 h a year. Before TPM: 24 failures, total repair time 240 h. After one year of autonomous + planned maintenance: 8 failures, total repair time 48 h.

• Before: MTBF = (7,200 − 240)/24 = 6,960/24 = 290 h; MTTR = 240/24 = 10 h; Ai = 290/(290+10) = 96.7%; λ = 1/290 ≈ 0.00345 /h.
• After: MTBF = (7,200 − 48)/8 = 7,152/8 = 894 h; MTTR = 48/8 = 6 h; Ai = 894/900 = 99.33%.
• MTBF improved 3.08 times; downtime fell from 240 h to 48 h = 192 h recovered. At ₹12,000 contribution per machine-hour, that is ₹23.04 lakh a year.
• Reliability over a 500-h order: before, R = e^−500/290 = e^−1.724 ≈ 0.178; after, R = e^−500/894 = e^−0.559 ≈ 0.572.

Example 4 — TEEP and the planned-downtime trap A line is loaded 2 shifts × 8 h × 25 days = 400 h/month out of a calendar 720 h; OEE = 70%.

• Utilisation = 400/720 = 55.6% → TEEP = 0.70 × 0.556 = 38.9%.
• If the plant reports OEE only, it looks 70% effective while 61% of calendar capacity never becomes good output (44% of it not even loaded). Adding a third shift at the same OEE would raise monthly good output by about 50% with no new machine — the classic "buy a machine or run the one you have" question.

Example 5 — comparing two machines at similar OEE

Same quality, similar OEE, opposite problems: X is available but slow/stopping often → focused improvement on minor stoppages and cycle time; Y is fast but down → planned maintenance, spares and SMED. Hence OEE must always be reported as its three factors, never as a single number.

Where it is used

• Automotive and auto components: Maruti Suzuki, Tata Motors, Bajaj Auto, Hero MotoCorp, TVS, Rane, Sundaram Clayton and Brakes India run pillar-based TPM; supplier OEE and PPM are audited as part of IATF 16949 customer requirements.
• Steel, cement and process plants: Tata Steel, JSW, UltraTech, Shree Cement — continuous process where one hour of kiln or mill downtime costs lakhs, so planned maintenance, vibration monitoring and shutdown planning dominate.
• Pharma and FMCG packaging lines: Cipla, Sun Pharma, HUL, Dabur, Nestlé India — changeover (setup) loss and start-up yield loss are the big ones, with GMP-compliant equipment logs doubling as maintenance records.
• Power, railways and heavy engineering: NTPC and state gencos (plant load factor and availability-based tariff), Indian Railways loco sheds and rolling-stock depots, BHEL and L&T shops.
• Job roles: maintenance engineer, TPM/CI (continuous improvement) coordinator, production engineer, reliability engineer, condition-monitoring analyst, quality engineer; typical fresher entry through GET/DET schemes where OEE improvement on a model machine is the first project.
• MSME clusters: cluster TPM programmes supported by productivity bodies (NPC, CII, QCFI) and MSME schemes, where a 10-point OEE gain usually removes the need for new capital equipment.

Interview questions

• Q: Define TPM in one line and state its goals. A: TPM is plant-wide productive maintenance carried out by all employees in small groups to maximise equipment effectiveness over its life, targeting zero breakdowns, zero defects and zero accidents.
• Q: What are the three factors of OEE and the six losses behind them? A: Availability (breakdown, setup/adjustment), performance (idling/minor stoppages, reduced speed) and quality (defects/rework, start-up loss); OEE is their product.
• Q: Why is planned downtime excluded from the OEE calculation? A: OEE measures how well the time the plant *intended* to produce was used; planned shutdowns are a scheduling decision, which is why TEEP is used to expose the capacity hidden in them.
• Q: Two machines both show 70% OEE — why ask more? A: Because the mix matters: a low-availability machine needs planned maintenance and faster changeovers, while a low-performance machine needs minor-stoppage and cycle-time kaizen.
• Q: How can OEE be inflated dishonestly? A: By padding planned downtime, using actual rather than design cycle time as "ideal", ignoring short stoppages, or counting reworked pieces as good.
• Q: Difference between TPM and RCM? A: TPM is a people- and loss-driven culture built on operator ownership and OEE; RCM is an analytical, FMEA-based method that chooses each maintenance task from its failure consequences — they complement each other.
• Q: What exactly do operators do in autonomous maintenance? A: Cleaning, lubrication, bolt tightening, visual inspection and simple adjustments against a standard, plus raising abnormality tags — overhauls and precision diagnostics stay with maintenance.
• Q: What is the difference between natural and accelerated deterioration? A: Natural deterioration is unavoidable wear in normal use; accelerated deterioration is extra wear from dirt, looseness, poor lubrication or wrong operation, and TPM removes it first because it costs nothing.
• Q: MTBF rose but availability hardly moved — why? A: MTTR or the waiting time for spares, permits and manpower is the binding constraint; availability needs both fewer failures and faster restoration.
• Q: Which pillar would you use for a chronic 2% dimensional rejection? A: Quality maintenance — identify the 4M conditions that produce the defect, set and monitor those equipment conditions and add poka-yoke, rather than inspecting more.
• Q: How does TPM support lean/JIT? A: JIT with low buffers collapses when machines stop, so stable, reliable equipment and fast changeovers from TPM are a precondition for flow and small lot sizes.
• Q: Which pillar covers Indian statutory compliance, and under what law now? A: The safety, health and environment pillar; statutory inspections of pressure vessels, lifting machines and safety devices come under the Occupational Safety, Health and Working Conditions Code, 2020 (in force from 21 November 2025) and state rules, and are never skipped for output.
• Q: First three things you would do on a line with 45% OEE? A: Measure losses for a week with real time-stamps, Pareto them into the six losses, then fix the top loss on one model machine with a cross-functional team while starting 5S and tagging.

Common mistakes

• Adding A, P and Q instead of multiplying them, or reporting a single OEE number without its three factors.
• Using the current achieved cycle time as the "ideal cycle time" — this hides all speed loss and makes performance look like 100%.
• Including planned downtime in loading time (deflating OEE), or dumping unplanned stoppages into "planned" (inflating it).
• Counting reworked or sorted pieces as good output, and forgetting that rejects also consumed machine time.
• Confusing MTBF with MTTF (repairable vs non-repairable items) and forgetting to subtract repair time from total time when computing MTBF.
• Treating TPM as a maintenance-department project with a 3-month deadline, instead of a 2-3 year management programme; or doing 5S and calling it TPM.
• In interviews, listing the eight pillars without being able to describe the 7 steps of Jishu Hozen or compute OEE from a shift log.

Related topics

Reliability engineering · Statistical quality control · Lean manufacturing and Six Sigma · ISO 9001 quality management systems · Maintenance planning and spare parts inventory · Condition monitoring and vibration analysis