Lean Maintenance3 mistakes ยท 4 fixes10 min read

5S in Maintenance(you're cleaning for audits, not for reliability)

The week before every audit, everyone cleans. Three weeks later, the workshop is back to baseline. 5S doesn't fail from lack of discipline โ€” it fails from the wrong posture.

10 min read
Lean Maintenance ยท 5S
3 mistakes ยท 4 fixes
By Melek Mehrez ยท June 29, 2026

The week before every audit, it's always the same. People clean, organize, label, repaint floor markings. The auditor validates. Three weeks later, the workbench is cluttered again. 5S in maintenance has a problem: it's used as an image tool, not a reliability tool.

The method comes from the Toyota Production System. Toyota didn't invent it so workshops look clean for photos. 5S has one objective: make problems visible before they become breakdowns. Without a visual reference for normal state, an oil leak, abnormal vibration, or worn cable goes unnoticed. With 5S properly applied, anomalies are immediately visible because they stand out against the standard.

That's where most teams miss: they treat 5S as a destination โ€” the tidy workshop โ€” rather than a means: early detection. This article diagnoses 3 recurring mistakes, reveals the blind spot nobody talks about (the 5S of CMMS data) and provides 4 concrete actions for a 5S that works for technicians, not auditors.

1

The three mistakes that make 5S fail

Stopping at S1 without sustaining S2

The first S โ€” eliminate the unnecessary โ€” is the easiest to do and hardest to maintain. In maintenance, this produces workbenches where "might be useful" parts accumulate, end-of-life tools kept "just in case", and manuals for machines decommissioned five years ago. You empty, you breathe. Then the volume returns, differently packaged.

Stopping there without doing S2 โ€” Seiton, giving each item a fixed place with a return rule โ€” produces the same result six months later.

What it costs โ€” SMRP data
15โ€“20%
of wrench time spent searching for tools or parts
Source: SMRP (Society for Maintenance & Reliability Professionals)
Not repairing
That time produces nothing โ€” it absorbs technician capacity that S2 can recover

Confusing Shine with "cleaning up"

In the Toyota spirit, S3 โ€” Seiso โ€” means inspecting while cleaning. You clean to see, not to look clean. When a technician cleans a machine with a Seiso protocol, they check seal condition, oil levels, unusual sounds, visible wear, and cable integrity. The cleaning is an inspection vehicle. In most workshops, Seiso is reduced to a mop pass. It looks clean. Nothing was inspected.

What TPS says

The Toyota Production System estimates a Seiso conducted with an inspection protocol detects an average of 30% of defects before they cause unplanned downtime. The result of a Seiso is measured by anomalies found โ€” not by the final visual.

Treating Sustain as audit preparation

Shitsuke โ€” maintain, discipline โ€” is the S nobody gets right. The structural reason: periodic 5S audits are scheduled (quarterly or annually) with individual motivation expected to carry it between them. Between audits, there's no signal, no indicator, no regular review. 5S becomes a matter of individual goodwill โ€” and goodwill always loses to daily operational pressure.

Direct consequence

5S is done for the audit, not because of it. The workshop is compliant during the visit and returns to baseline within three weeks. The cycle repeats as the next audit approaches.

2

The blind spot nobody talks about: 5S for your data

Physical 5S and digital chaos coexist comfortably

In most facilities: immaculate Kanban shelves โ€” and a parts catalog with 400 duplicate referencesbecause the same part was entered under five different names by five different technicians. A perfect shadow board โ€” and equipment in the CMMS named "Pump P1", "Pump 1", "PUMP-01", "P01" depending on who did the entry. Clear aisles โ€” and maintenance histories full of "Repaired OK" with no root cause.

What happened with KPIs that lie and ghost maintenance has the same root: unusable data because it was never subjected to the equivalent of 5S.

Physical 5SCMMS data equivalent
S1 โ€” Sort: eliminate the unnecessaryDelete duplicate part references, ghost equipment, WOs closed with no information
S2 โ€” Set in order: a place for everythingStandardize nomenclature: equipment hierarchy, ISO 14224 failure codes
S3 โ€” Shine: clean to inspectRegularly audit data quality (% WOs with root cause, % empty fields)
S4 โ€” Standardize visual rulesClear entry procedures, mandatory fields at WO closure
S5 โ€” Sustain over timeData quality dashboard visible in team review, updated every week
Why this matters: without data 5S, reliability analysis produces inconsistent results month to month โ€” not because equipment behavior changes, but because the same failure is coded differently depending on which technician closes the WO. You lose exactly what you were trying to gain.
3

4 actions for a 5S that works for technicians

1

Start with spare parts stock, not the workshop in general

The most profitable first 5S project in maintenance isn't the workshop as a whole โ€” it's the spare parts stock. Three questions per reference: does it correspond to equipment still in service? What is its shelf life (some electronics degrade, seals dry out)? Is it a duplicate?

Typical result: 20 to 35% obsolete or duplicate in an average-size stock โ€” capital tied up for nothing.
2

Add kitting to your S2

Shadow boards work well for common tools. But in industrial maintenance, there's a more profitable dimension: kitting. Preparing in advance โ€” in a labeled bin โ€” all the tools, parts, and consumables needed for a planned intervention, before the technician goes to the field. A technician who makes three store trips during an intervention loses time and multiplies reassembly error risk. S2 done right: every recurring intervention type has its standard kit ready the day before.

3

Turn your cleans into inspections

Create a visually simple checklist โ€” A4, laminated, attached to the machine โ€” with points to check during Seiso: seal condition, oil level, abnormal sound, unexpected temperature, visible cable integrity. The cleaning follows the inspection protocol.

In the CMMS: an inspection form per equipment โ€” checkboxes, flagged anomaly โ†’ work order created automatically.
4

Replace annual audits with weekly mini-audits

A timestamped "standard state" photo displayed per zone. A 5-minute visual mini-audit each week (10-point checklist, zone owner). A compliance indicator visible in the workshop, updated weekly. 5S integrated into operational performance meetings โ€” Daily or Weekly โ€” instead of reserved for the annual visit. What gets measured regularly sticks. What only gets measured at audit time lasts until the audit.

One-Point Lessons (OPL) โ€” S4 in maintenance

S4 โ€” Standardize โ€” takes the form of One-Point Lessons: one page, one visual, one operating rule or one anomaly sign to know. Format: A4, photo or diagram, three lines maximum. Attached to the machine or integrated into the equipment record in the CMMS, an OPL eliminates dependence on tribal knowledge: the replacement technician sees the same standards as the regular.

Measuring the 5S that builds reliability (not the one that impresses)

The simplest test: how many anomalies did you find during your last Seiso? If the answer is zero, you did cosmetics.

IndicatorCosmetic 5SReliability 5S (target D+90)
Anomalies found in Seiso / week0 โ€” "everything's fine"3 to 5 per equipment visited
Tool/part search time (min/intervention)15 to 25 minUnder 8 min
% planned interventions with pre-prepared kitUnder 20%Over 70%
% WOs with root cause filled inUnder 40%Over 80%
Duplicate part references eliminatedBaselineโˆ’ 30% or more
5S audit frequencyAnnual (before the visit)Weekly (5 min, integrated into review)

Further reading

  • Toyota Production System (Taiichi Ohno, 1978) โ€” origin and objective of 5S as a problem-visibility tool
  • SMRP โ€” maintenance productivity metrics (wrench time, typically 25 to 35% of total technician time)
  • ISO 14224:2016 โ€” failure classification and cause codes (basis of CMMS data 5S)
  • Article #3 in the series โ€” Ghost Maintenance (tribal knowledge, uncaptured data โ€” data 5S closes the same leaks)

A 5S that detects failures needs structured inspection rounds

FreeMaint is a freemium CMMS (Core free forever). Equipment inspection rounds, stock management with threshold alerts, and failure history are included โ€” so Seiso produces data, not photos for the audit.

Article by Melek Mehrez, founder of FreeMaint. All articles.

Frequently asked questions

Why does 5S never stick in maintenance?

5S doesn't stick because it's only measured during periodic audits โ€” with individual motivation expected to carry it between them. The fix: replace the annual audit with a weekly 5-minute visual mini-audit (10-point checklist, zone owner). What gets measured regularly sticks. What only gets measured at audit time lasts until the audit.

What does Seiso (S3) actually mean in maintenance?

In the Toyota spirit, Seiso means 'clean to inspect', not 'clean up'. A Seiso protocol includes checking seal condition, oil levels, unusual sounds, visible wear, and cable integrity. The result is measured by anomalies found, not by how clean things look.

What is kitting in maintenance?

Kitting means preparing in advance โ€” in a labeled bin โ€” all the tools, parts, and consumables needed for a planned intervention, before the technician goes to the field. In properly applied 5S, each type of recurring intervention has its standard kit ready the day before.

How do you apply 5S to CMMS data?

Data 5S follows the same logic: S1 โ€” eliminate duplicate references; S2 โ€” standardize nomenclature (ISO 14224); S3 โ€” audit data quality regularly; S4 โ€” mandatory fields at WO closure; S5 โ€” a data quality dashboard in team review every week.