5S Sustainment Audit: How to Tell If Workplace Order Is Slipping Before It Collapses

The 5S Sustainment Audit audits the 5S discipline in a work area to sustain workplace order. You score each of the five elements, Sort, Set in Order, Shine, Standardise and Sustain, against a stated criterion, record what you found, and name an action where the score falls short.

It sits in the Sustain phase of RDMAICS for a reason. By the time you reach Sustain, the improvement has been handed over. The project team has moved on. The visual layout, the labelling and the standard work are now owned by the people doing the job every day. The question is no longer whether the change works. The question is whether it is still there in three months.

Use the tool on a schedule, not on a hunch. A monthly or fortnightly audit of the same area with the same criteria gives you a trend. A single audit gives you a snapshot, which is useful but weaker. The value compounds when you run it repeatedly and watch which element keeps losing a point.

Also use it when something feels off. If a team starts missing its board update, or a workstation looks cluttered, the audit converts that feeling into a score and a finding you can act on.

Everything below is built on the sample data that loads with the tool. The case is fictional. It is there to show the mechanics, not to describe any real site.

The audit covers one work area with the five standard elements. Each element carries a maximum score of 5, so the audit has a ceiling of 25. Three elements score 4. Two score 5. Each element has a finding recorded against it, and three of those findings carry an action.

This is the pattern you will see most often in a healthy area some months after handover. Nothing has failed. Nothing is dramatic. But small deductions are appearing in a few places, and the pattern of where they appear tells you more than the total.

The input table has one row per 5S element. For each row you supply the element name, the criteria you are scoring against, the maximum score, the score you awarded, the finding, and the action if one is needed. Element, max score, score and finding are required. Action is optional, because a conforming element does not need one.

The criteria column is the part people rush, and it is the part that decides whether the audit is repeatable. In the sample data the criteria are short and testable. Sort is scored against "Only needed items present". Set in Order against "Items organised and labelled". Shine against "Workspace clean". Standardise against "Standards documented". Sustain against "Discipline maintained".

Short criteria work because two different auditors can read them the same way. If your criteria need a paragraph to explain, your scores will drift between auditors and your trend will be noise.

The findings in the sample data follow the same discipline. They are specific and observable. "1 obsolete doc". "Dusty scanner". "Board 1 day late". "Conforming" where nothing was wrong. None of them are opinions. Each one could be verified by a second person walking the same area.

The tool calculates a total score of 22 of 25, a score as a percentage of 88 per cent, an average score of 4.4 out of 5, elements at full score of 2 of 5, and findings with an action of 3.

Read these in order of usefulness, and the order is not the order they appear. The total and the percentage are the headline, but they are the least actionable numbers on the screen. Eighty eight per cent sounds settled. It is the other three figures that carry the signal.

Elements at full score is 2 of 5. That means three of the five elements have an open gap. Findings with an action is 3, which matches. So the area is not holding at full standard in most of what you audit. The average of 4.4 out of 5 confirms the deductions are shallow rather than deep. Nothing scored 2 or 3. The slippage is consistent and small, not concentrated and severe.

The tool also lists each element as a card showing the finding, the action, the criteria and the maximum score, so you can go straight from the number to the thing you saw. It offers an AI review that explains the result as well, if you want a written reading of the same data.

Sort scored 4 of 5 with the finding "1 obsolete doc" and the action Remove. Shine scored 4 of 5 with the finding "Dusty scanner" and the action Clean. Sustain scored 4 of 5 with the finding "Board 1 day late" and the action Coach.

Set in Order and Standardise both scored 5 of 5 and were recorded as Conforming. That matters. The structural work has held. The labelling is in place. The standards are documented. Nobody has dismantled the system.

What has slipped is the routine around the system. An obsolete document was not removed. A surface was not wiped. A board update ran a day late. These are habit failures, not design failures. The fix for a habit failure is different from the fix for a design failure, which is why the action on Sustain is Coach rather than a rework of the standard.

That is the early warning you are looking for after handover. When Set in Order and Standardise stay at 5 while Sort, Shine and Sustain start dropping to 4, the system is intact and the discipline is thinning. Catch it there and a conversation fixes it. Leave it and the structure starts to follow.

The first decision is whether to escalate. With an average of 4.4 out of 5 and no element below 4, you do not escalate. You assign the three actions and re audit. If an element had dropped to 2 or 3, the answer would change.

The second decision is where to spend coaching time. The Sustain element scored 4 with the action Coach. That is a people decision, not a housekeeping one. It points you at the board owner and the routine, not at the work area.

The third decision is what to do with the actions themselves. Remove, Clean and Coach are three different kinds of work. Remove and Clean are one off tasks with a clear end. Coach is ongoing. Tracking them as one undifferentiated list of three is how actions quietly stay open.

The fourth decision is whether to change the criteria. If an element sits at 5 of 5 for many audits in a row, your criterion may be too easy, or the risk may genuinely have gone. Both are worth knowing before you keep auditing it.

The most common mistake is treating the percentage as the result. Eighty eight per cent is a summary of five judgements. On its own it hides that three of five elements have open gaps. Teams that chase the percentage start scoring generously, and the audit stops detecting anything.

The second mistake is scoring an element down without recording a finding. A score of 4 with no finding is unusable. Nobody can act on it and nobody can verify it next month. Every deduction in the sample data has a matching finding and a matching action.

The third mistake is a vague action. "Improve" or "Review" are not actions. Remove, Clean and Coach are, because each one names what someone does.

The fourth mistake is changing the criteria between audits without saying so. Your trend then compares two different measurements. If you must change a criterion, note the change and treat the trend as restarting.

The audit does not measure whether 5S is worth doing in this area. It measures whether the 5S standard you set is being kept. If the layout was wrong at design time, a score of 22 of 25 tells you the wrong layout is being maintained well.

It does not weight the elements. Sort, Shine and Sustain each lost one point, and each loss counts the same in the total. In practice a late board update and a dusty scanner may carry very different risk. The tool will not make that judgement for you.

It does not verify that the actions were completed. Findings with an action is 3, and that number stays at 3 whether the work happens or not. Closure has to be tracked outside this audit, and confirmed at the next one.

It does not produce a trend by itself from one run. A single result of 88 per cent is a point, not a direction. You need repeated audits with stable criteria before you can say anything about whether the area is improving or slipping.

Finally, it does not replace being in the area. The scores are only as good as the walk that produced them. The tool organises what you saw. It does not see for you.