Total Productive Maintenance (TPM) is a Lean approach to equipment management in which operators, maintenance technicians, engineers, and managers share responsibility for keeping machines running at full capability. Its goals are zero breakdowns, zero defects, and zero accidents. The "total" means every piece of equipment, every person who touches it, and the whole equipment lifecycle from specification to retirement. The practical shift is that routine care (cleaning, inspection, lubrication, minor adjustment) moves to the operators who run the machine, and the maintenance department's time is freed for planned work, diagnosis, and improvement.
Where TPM came from
TPM was developed in Japan in the 1960s and 1970s as an extension of the American preventive maintenance practices that Japanese industry had adopted after World War II. Nippondenso (now Denso), a Toyota group supplier, is credited with the first full implementation in 1971, when its automated lines made it impractical for a separate maintenance department to keep up and operators were trained to maintain their own equipment. Seiichi Nakajima of the Japan Institute of Plant Maintenance (JIPM) codified the approach, defined the pillars, and introduced it to the West with Introduction to TPM (English edition 1988). JIPM still administers the TPM Excellence Award, which sets the reference standard for what a mature program looks like.
The 5S foundation
TPM is built on 5S, for a specific reason: you cannot inspect a machine you cannot see. A press covered in chips and oil hides a loose bolt, a cracked guard, and a leaking fitting equally well. Sort and Set in Order clear the area so the equipment is accessible; Shine is where TPM begins, because cleaning is inspection, and the operator who cleans a machine daily is the first to notice something changing; Standardize turns that cleaning into a checklist with a frequency; Sustain is the audit that keeps it running. Facilities that try to launch TPM without a working 5S program usually end up doing 5S first anyway.
The eight pillars of TPM
JIPM's eight pillars are the activities a TPM program is organized around. Most programs start with the first two and the fourth, and add the rest as they mature.
1. Autonomous maintenance (Jishu Hozen)
Operators take over routine care of their equipment: daily cleaning, inspection against a checklist, lubrication, tightening, and simple adjustments. They are trained to recognize abnormal conditions (noise, vibration, heat, leaks, loose fasteners) and to tag them for maintenance. JIPM defines seven steps for rolling it out, from initial deep cleaning through operator-set standards to full self-management. The point is not to turn operators into technicians; it is to catch deterioration at the stage where it costs minutes instead of days.
2. Planned maintenance
The maintenance department moves from reacting to breakdowns to scheduled work based on equipment history, manufacturer guidance, and condition monitoring: preventive maintenance on a calendar or run-hour interval, predictive maintenance using vibration, thermal, or oil analysis, and a parts and spares plan so planned work is not delayed waiting for components. As autonomous maintenance takes over the routine tasks, planned maintenance gets the time to do this.
3. Quality maintenance (Hinshitsu Hozen)
Identifies the equipment conditions that produce defects and controls them, so quality is built in by the machine's state rather than inspected out. If bearing wear on a spindle produces out-of-tolerance parts, quality maintenance sets a bearing inspection and replacement standard tied to the tolerance, not just to the bearing's failure.
4. Focused improvement (Kobetsu Kaizen)
Cross-functional teams attack the largest equipment losses one at a time, using root cause analysis, Pareto analysis, and kaizen events. The six big losses below are the usual targets.
5. Early equipment management
Applies what maintenance and operations have learned to the specification, design, and installation of new equipment, so new machines arrive easier to clean, inspect, and maintain, and reach full output faster. Maintenance and operators review equipment purchases before they are made.
6. Training and education
Operators learn the inspection and maintenance skills autonomous maintenance requires; technicians learn diagnosis and condition monitoring; managers learn to coach rather than direct. Most programs keep a skills matrix per area showing who is qualified for what.
7. Safety, health, and environment
Zero accidents is one of TPM's three goals. Every pillar is applied with the question of what could injure someone: lockout before inspection, guards restored after cleaning, hazards found during autonomous maintenance tagged and fixed. This is also where TPM and 6S overlap.
8. TPM in administration
Applies the same loss-elimination thinking to the office processes that feed the floor: order entry, scheduling, purchasing, and the maintenance planning function itself. A machine waiting on a purchase order for a part is a loss the floor cannot fix.
The six big losses
Nakajima defined six categories of equipment loss. They map directly to the three factors in OEE and are what focused improvement teams work on.
| OEE factor | Loss | Examples |
|---|---|---|
| Availability | Breakdowns | Unplanned stops from equipment failure |
| Setup and adjustment | Changeovers, tooling changes, warm-up, adjustment after a changeover | |
| Performance | Minor stops and idling | Jams, sensor blocks, misfeeds, waiting on material; stops short enough that nobody logs them |
| Reduced speed | Running below the designed cycle time to avoid defects or stops | |
| Quality | Defects and rework | Parts made out of spec during normal running |
| Startup losses | Scrap produced between startup and stable running |
Measuring TPM: Overall Equipment Effectiveness
OEE is the single metric TPM programs track. It expresses what fraction of planned production time a machine spent making good parts at its designed rate.
OEE = Availability × Performance × Quality
- Availability = Run time ÷ Planned production time. Captures breakdowns and setup losses.
- Performance = (Ideal cycle time × Total count) ÷ Run time. Captures minor stops and reduced speed.
- Quality = Good count ÷ Total count. Captures defects and startup losses.
A line planned for 480 minutes that ran 400 (availability 83%), produced 3,200 parts in those 400 minutes against an ideal of 10 per minute (performance 80%), with 3,040 good (quality 95%) has an OEE of 0.83 × 0.80 × 0.95 = 63%. World-class OEE is often quoted as 85%, but the useful number is the trend for one machine over time and the breakdown that shows which of the three factors is costing the most. OEE is a diagnostic, not a scorecard to compare plants against each other.
How to start a TPM program
- Secure management commitment and announce it. TPM changes job roles, which does not happen without leadership saying so and funding the training time.
- Pick a pilot machine. One that matters (a bottleneck or a chronic problem), that operators care about, and where improvement will be visible. Not the worst machine in the plant; that one is demoralizing.
- Baseline OEE. Measure the pilot machine for two to four weeks before changing anything, and log the six losses so the team knows which to attack.
- Do 5S on the pilot area, including an initial deep clean of the machine with operators and maintenance together. This clean usually surfaces a list of abnormalities (loose, worn, leaking, missing) that becomes the first work orders.
- Launch autonomous maintenance. Operators and maintenance write the daily and weekly inspection checklist together, with photos and visual standards at the machine: lubrication points color-coded and labeled, gauge ranges marked, inspection points numbered. Train operators on each item and on what abnormal looks like.
- Run the first focused improvement. Take the largest loss from the baseline through root cause analysis and fix it. Re-measure OEE and post the result at the machine.
- Build the planned maintenance schedule for the pilot from equipment history and the abnormality list, and move routine tasks off the maintenance department's list now that operators own them.
- Standardize and expand. Document what worked on the pilot, then take the next machine with the pilot team as coaches. Add the remaining pillars as the program matures.
Benefits
- Fewer breakdowns, because deterioration is caught daily instead of at failure.
- Higher OEE on the same equipment, which is capacity without capital spending.
- Fewer defects, because equipment condition is controlled rather than compensated for.
- Lower maintenance cost over time: planned work costs less than emergency work, and parts are replaced on condition rather than on failure.
- Fewer injuries, because operators know their equipment and abnormal conditions are fixed rather than worked around.
- Operators who understand and take ownership of their machines, which is the change that makes the other benefits stick.
Common reasons TPM programs fail
- Launching autonomous maintenance without giving operators time in the schedule to do it.
- Treating it as a maintenance department initiative rather than an operations one.
- Skipping the 5S foundation and trying to inspect dirty, inaccessible equipment.
- Measuring OEE and posting it without a focused improvement process to act on it.
- Rolling out plant-wide instead of proving it on a pilot machine first.
Frequently asked questions
What is the difference between TPM and preventive maintenance?
Preventive maintenance is scheduled maintenance done by the maintenance department. TPM is a plant-wide program in which operators do routine care, maintenance does planned and predictive work, and cross-functional teams eliminate equipment losses. Preventive maintenance is one pillar of TPM.
What are the three goals of TPM?
Zero breakdowns, zero defects, and zero accidents.
What is a good OEE?
85% is commonly cited as world class, and 60% is typical for a plant that has not measured it before. The more useful number is your own trend and the breakdown by availability, performance, and quality.
Is TPM part of Lean?
Yes. It is the Lean approach to equipment, developed in the Toyota group alongside the rest of the Toyota Production System. Unreliable equipment makes flow, pull, and kanban impossible, which is why TPM is usually implemented early in a Lean program.
How long does TPM take to implement?
A pilot machine shows measurable OEE improvement in two to three months. A plant-wide program with all eight pillars is a multi-year effort; JIPM award-level programs typically take three to five years.
Related resources
TPM sits alongside 5S, kaizen, kanban, and root cause analysis in the Lean toolkit, and overlaps with quality control on the quality maintenance pillar. The Lean Enterprise Institute's lexicon entry on TPM is a concise reference. 5S Today stocks the labels and visual standards and floor marking that autonomous maintenance checklists are built on.