Total Productive Maintenance has a branding problem: the textbook version arrives as an eight-pillar program with committees, master plans, and a multi-year horizon. Most Swiss plants that need TPM do not need that. They need one thing first: equipment that runs when the plan says it should. That is buyable as a scoped project.
The honest starting point: measure OEE without flattering yourself
Overall Equipment Effectiveness - availability × performance × quality - is only useful if the losses are recorded honestly. The first two weeks of a scoped TPM start are usually spent making the number true: logging minor stops that never reached any system, timing changeovers as they really run, counting startup scrap. It is common for a "we run at 78%" line to measure at 55 to 65% once minor stops are honest. This is good news; it is recoverable capacity you have already paid for.
The 12 to 16 week scoped format
Weeks 1 to 2 - loss picture. Honest OEE on one constrained line, Pareto of the top losses. The top three losses set the scope; everything else waits.
Weeks 3 to 6 - restore basic conditions. Deep clean as inspection, tagging every defect found, fixing the backlog. A line that has drifted for years typically surfaces 100 to 300 tags. This phase overlaps heavily with 5S - if the area has no workplace organisation, do that first or simultaneously.
Weeks 7 to 12 - autonomous maintenance. Operators take over cleaning-inspection-lubrication routines with visual standards: check points marked on the machine, quantities and intervals photographed. Maintenance stops being the janitor of neglect and gets time back for real planned work.
Weeks 12 to 16 - one planned-maintenance loop. Failure data from the tags feeds a simple planned schedule for the components that actually fail. Not a CMMS project - one loop that demonstrably prevents repeat failures, plus a weekly equipment review where production and maintenance look at the same numbers.
Typical outcome on a constrained line: 10 to 20 OEE points recovered, most of it from minor stops and changeover-adjacent losses. On equipment that gates plant output, the payback math is rarely subtle.
What to defer (possibly forever)
Early-equipment-management pillars, plant-wide TPM offices, certification audits: defer until the first line has held its gains for two quarters. The eight pillars describe a mature state - they are a terrible to-do list for month one. A practitioner who opens with a pillar assessment instead of a loss Pareto is running a program, not fixing your equipment.
FAQ
How long does a TPM implementation take?
A scoped start on one line takes 12 to 16 weeks to honest OEE, restored basic conditions, running autonomous maintenance, and one planned-maintenance loop. Plant-wide maturity is a multi-year path - but you buy it line by line, funded by each line's gains.
What OEE improvement is realistic?
On a line that has never had structured loss elimination, recovering 10 to 20 OEE points within the first months is typical, dominated by minor stops, changeover losses, and startup scrap.
What is the difference between TPM and preventive maintenance?
Preventive maintenance is a maintenance-department schedule. TPM moves routine care (cleaning, inspection, lubrication) to operators, uses real failure data to focus planned work, and makes equipment losses a shared production-maintenance number. The cultural shift is the point; the schedule is a component.
Do we need 5S before TPM?
Practically, yes for the pilot area: autonomous maintenance standards do not survive in a disorganised workplace. The restore-basic-conditions phase can carry a compact 5S within it if the area is close.