TPM — Total Productive Maintenance — is a philosophy, or a tool for production management, depending on the scope you apply it at. Its ultimate goal is improved company performance, which shows up as product quality, cost reduction and control, on-time delivery, and better safety and environment.
Equipment is one of the most important factors in making quality products at low cost and on time. But managing equipment effectiveness is usually obstructed by how management, production and maintenance actually work together — which is the problem TPM was created in Japan to solve, and which carried many Japanese manufacturers to world-class manufacturing.
TPM is a philosophy, not a project with an end date
TPM is used in two quite different senses. The first is a set of production-management tools — forms, boards, meetings. The second is a way of looking at equipment and at the people who run it. A plant that takes the first meaning will have a complete document system within a year and abandon it in the second. A plant that takes the second changes how the floor thinks, permanently.
The stated goal of TPM is very short: zero breakdown, zero defect, zero accident. Those three zeros are not there to be reached; they are there to set the direction. A plant that targets a ten per cent reduction stops thinking at ten per cent. A plant that targets zero has to ask a different question — what allows this to happen at all.
Equipment is the least-managed factor in the plant
Every plant manages its people, its materials and its costs seriously — with numbers, an owner and a monthly meeting. Equipment effectiveness is usually the exception, because a machine that is still running is assumed to be fine, even when it has been running twenty per cent below its design rate for three years.
Equipment loss is therefore invisible in the accounts. It does not appear as a cost line; it appears as capacity that is not there, as overtime paid to catch up with the plan, and as capital spent on a new machine while the existing one has never been used to its capability. TPM begins by making that loss visible, and only then by attacking it.
Two phases: production first, then the whole company
TPM runs in two phases that have to be taken in order. The first is Production TPM, applied inside production only, built on five pillars and aimed at the three zeros. When that phase is genuinely working, it extends into Company-wide TPM, which adds three more pillars and reaches every operating unit — office, sales, purchasing included.
That order matters more than most people expect. Plants that declare company-wide TPM on day one usually end up with a TPM board on every department wall and no department reducing any loss, because nobody has yet seen a tangible success in production to copy.
The eight pillars of TPM
- Individual improvement
- Autonomous maintenance
- Planned maintenance
- Education and training to raise operating and maintenance skill (operation and maintenance development)
- Maintenance considered from the design stage (initial phase management)
- Quality maintenance
- A way of working for the administrative departments that is alive to production efficiency, called TPM in the office
- Health, safety and the working environment within the plant
Although most TPM activity must be carried out by the production and maintenance departments, it is a mistaken idea to have those two departments do all of it. The pillars of TPM should be run as small-group activity whose members come from every department concerned, according to the nature of the pillar in question.
The pillars, described as they are actually used
Production carries five pillars. The first is focused improvement: a cross-functional team picks a loss that can be counted and attacks it until it is gone. This is the engine of the whole system. The second is autonomous maintenance, where operators hold the basic condition of their own machines — cleaning, lubricating, adjusting, inspecting and minor repair. The third is planned maintenance, where the maintenance department moves from reacting to working to a plan set by criticality.
The fourth is skill development, for operators and technicians alike, worked through a per-person, per-machine skill map. The fifth is initial-phase management: feeding what the plant learned from its existing equipment into the specification for the next machine, before it is ordered rather than after it is installed.
Extending to eight adds quality maintenance, which traces each defect type back to the equipment condition that produces it; office improvement, which deals with the paperwork that keeps production waiting; and safety, health and environment — not a pillar added at the end, but a condition that has to hold from the first day.
How the eight pillars group
Pillars 1, 2 and 3 are the ones that have to be carried out to bring TPM about in production. Before that work starts and while it runs, training and skill development must go on continuously, and that is the duty of pillar 4. Pillar 5 is the advanced stage of TPM in production, because it plants maintenance into the equipment itself, into the material, the production method, the way the work is done, and into the design and layout of the plant or the process. Pillars 6, 7 and 8 are carried out to extend TPM from production to TPM across the whole organisation.
Although most TPM activity has to be carried out by production and maintenance, it is a mistake to leave the whole of it to those two departments. The pillars of TPM should be run as small-group activity, with members drawn from every department the pillar concerns.
Twelve steps, in four phases
Deployment follows twelve standard steps in four phases. Preparation is where the most senior manager announces the policy personally, everyone at every level is educated, the committees and working teams are formed, measurable policy and targets are set, and the master plan is drawn up. It is the longest phase and the one most often rushed.
Kick-off is the formal launch — not an empty ceremony, but the moment everyone in the plant hears the same thing at the same time. Then comes implementation, the real pillar-by-pillar work, and finally consolidation: audit, certification, and a new set of targets pitched higher than the last.
Overlapping small groups
The structure that actually makes TPM work is a set of small groups that overlap: the leader of a group at one level is a member of the group above it. Targets from management reach the floor without being lost, and problems from the floor reach management without being filtered out on the way.
Where groups are set up in parallel rather than overlapping, each one works hard in its own direction and nobody can say where the sum of that effort is taking the plant.
Why TPM programmes fail
The most common cause is not technical: it is management handing ownership of TPM to the maintenance department. Once that happens, production treats it as somebody else’s programme, and autonomous maintenance never starts — because the people who have to do it are operators, and operators belong to production.
The next is measurement that stops. A plant collects OEE data intensively for three months and lets it lapse in the fourth. Without data the meeting becomes a report of impressions, and once it does, the busiest people stop attending first.
The third is expecting results too soon. Restoring the basic condition of equipment neglected for ten years takes months, not weeks, and the numbers often get worse first, because abnormalities that were always there are being seen and recorded for the first time.
What TPM is aiming at
Developing the equipment means mobilising everyone to take part, so as to produce the highest equipment effectiveness, reliability, product quality, machine productivity, and the ability to use the machine across its whole service life.
Developing people means putting each department in a position to carry its own part of TPM:
- Operators can maintain their own machines
- Production can reduce the loss the equipment causes
- Maintenance can maintain the equipment to a plan
- Design, research and development work with maintenance in mind from the start
- Everyone can work together effectively as small-group activity
And then the organisation
Out of developing the equipment and developing the people comes the development of the organisation, in the form of productivity improvement, quality improvement, cost reduction, delivery on time, safety, morale, and care of the environment.
Raising overall equipment effectiveness to its highest by strengthening the relation between the operator and the machine — as though the operator owned it — is the goal of TPM. And developing people who know how to work as a team is the walk towards that goal.
The operational indices
The indices of operational success divide by level — plant, division, department, and the autonomous maintenance group.
| Level | Indices |
|---|---|
| Plant | Production cost · labour cost · value added to material · labour productivity rate · work in process · customer complaints · number of accidents |
| Division | Overall equipment effectiveness (OEE) · time productivity rate · labour productivity rate · production cost · cutting tool cost · material cost · overtime · labour saved · customer complaints · maintenance cost · jig and fixture cost · energy cost · accidents · results of focused improvement · material utilisation rate · number of suggestions |
| Department | Number of machine stoppages · reduction in setup time · number of breakdowns · number of cutting tool changes · start-up period · number of minor stoppages · number of speed losses · number of defects · reduction in production time · increase in production speed · yield rate · labour saved · reduction in labour hours · number of suggestions |
| AM group | Reduction in cleaning time · reduction in inspection and lubrication time · reduction in oil consumption · number of points improved · number of abnormalities found · number of causes traced · reduction in minor stoppages · reduction in setup time · reduction in tool change time |
The organisational performance indices
Measures for production, quality, cost, delivery, and then safety and the morale of the organisation’s people.
| Area | Measure | How it is measured |
|---|---|---|
| Production | Productivity | Output / input |
| Production | OEE · operating rate · operating efficiency · quality rate | Overall equipment effectiveness measurement |
| Production | MTBF | Mean time the machine runs without failing |
| Production | MTTR | Mean time the machine is down when it fails |
| Production | Return on equipment investment | (Value produced / value of equipment) x 100 |
| Production | Planned maintenance rate | (Maintenance done / maintenance due) x 100 |
| Production | Preventive maintenance rate | [(PM + CM) / (PM + CM + BM)] x 100, by cost |
| Production | Material utilisation rate | (All material entering the process / material leaving as product) x 100 |
| Quality | Defects · customer complaints · returns | From the record |
| Quality | Quality rate | Overall equipment effectiveness measurement |
| Cost | Production cost · maintenance cost · value of equipment retired | From the record |
| Cost | Capital turnover | Net sales / fixed assets |
| Cost | Depreciation rate | (Depreciation / fixed assets) x 100 |
| Delivery | Late deliveries · days late · products delivered late | From the record |
| Safety | Accident rate | (Time lost to accidents / hours worked) |
| Safety | Accident-free working time | (Working days with an accident / total working days) |
| Morale | Number of group activities · time spent on them · number of suggestions · training hours | From the record |
| Morale | Autonomous maintenance rate | (Repairs made by the operator / all breakdowns) x 100 |
Why measure at all
The only thing that will tell us how TPM is going, whether it is on the right road, and how much it is achieving, is measurement at intervals — to know the progress, to stop and review, or to adjust the plan for what comes next.
How to apply it on the floor
- Start on one model line before rolling out plant-wide
- Have senior management declare the policy themselves, not HR on their behalf
- Name the committee and the pillar leaders in the first month
- Set targets as measurable numbers, not ‘improve things’
- Write a master plan that says who does what, by when
The mistakes we see most
- Launching every line at once with the same resources, which produces a shallow implementation everywhere
- Handing TPM to maintenance alone, when six of the eight pillars sit outside that department
- Running TPM as a project with an end date rather than as the way the plant works
- Measuring the number of activities held instead of the loss actually removed
Common questions
How long before TPM shows results?
Working on a model line and attacking the largest loss first, measurable change usually appears within three to six months. Making the culture stick takes eighteen months or more.
Can a small plant do TPM?
Yes, and often more easily than a large one, because the chain of command is shorter. What it needs is a serious management sponsor and someone clearly tasked with driving it.