Key takeaways
  • Sort, Set in order, Shine, Standardize, Sustain. The first four are a project; the fifth is a habit, and it is where almost every programme fails.
  • The common mistake is treating 5S as a tidy-up. It is a system for making abnormal conditions visible at a glance.
  • Sustain needs a scheduled audit with an owner and a score, not a poster.
Where 5S programmes die
Month three, when the launch team stops walking the floor

Sort, Set, Shine and Standardize can all be done in a fortnight of effort. Sustain is the only one that requires something to happen every week forever, which is why it is the step that quietly gets dropped.

What the five S's actually mean

  • Sort (Seiri). Remove everything not needed for the work being done here. Red-tag anything uncertain and hold it for thirty days rather than debating it.
  • Set in order (Seiton). Give everything that remains a defined home, chosen by frequency of use and reach, and label it so a stranger could return it correctly.
  • Shine (Seiso). Clean as inspection. The point is not appearance, it is that a leak, a crack or a loose fastening is visible when the surface underneath is clean.
  • Standardize (Seiketsu). Write down what good looks like for this area, with photographs, so the standard survives the person who set it.
  • Sustain (Shitsuke). Audit against that standard on a schedule, with a named owner and a score that is visible.

The mistake that makes it a tidy-up

If the visible outcome is a cleaner area, you have done housekeeping. The outcome that matters is that an abnormal condition becomes obvious: a missing tool leaves a marked gap, a leak shows on a clean floor, a level outside range is read at a glance rather than calculated.

A useful test after Set in order: could someone from another line find the right tool in ten seconds, and would they notice if it were missing? If not, the area is tidy but not organised.

Doing Sort without an argument

Disputes about whether something is needed consume more time than the sorting. Red-tag anything contested, move it to a holding area with a date, and review after thirty days. Almost nothing is reclaimed, and the decision is made by evidence rather than by whoever argues hardest.

How to make Sustain real

  • Audit weekly at first, then monthly once scores stabilise. Fortnightly is a good long-term rhythm for most areas.
  • Score against the photographed standard, not against an inspector's opinion.
  • Give each area a named owner who walks it, rather than a central 5S coordinator who walks everywhere.
  • Make findings into tracked actions with owners and dates. An audit that produces a score but no work is theatre.
  • Publish scores where the team works, not in a management pack.

A realistic first month

Pick one area, ideally one that frustrates the people in it. Week one, sort and red-tag. Week two, set homes and label. Week three, deep clean and photograph the standard. Week four, run the first audit and fix what it finds. Then repeat in the next area while auditing the first, which is the point at which you discover whether Sustain is real.

Keeping the audit from lapsing

The practical failure is that the audit is somebody's intention rather than something scheduled. Whatever system you use, the audit should arrive at the owner on a schedule, record a score with photographs, and turn findings into tracked work. In RakuOps that audit can be sent to the area owner as a link, completed on the floor, and escalated if it is missed, so Sustain does not depend on anyone remembering.