5S Workplace Design: Turn a Diagnosed Cause Into a Pilot You Can Size

5S Workplace Design asks you to apply five pillars to a workspace. Sort, Set in Order, Shine, Standardise and Sustain. The output is a designed future state rather than a tidy-up. You describe the current state for each pillar, name the action that moves it, and attach an owner and a target date.

The tool belongs in the Improve phase of RDMAICS. By the time you open it, Analyse should already have told you which cause is doing the damage. 5S is one of the responses to that cause. It fits when the cause is physical or environmental. Clutter, scattered tooling, inconsistent housekeeping, methods that differ by person, and no routine to hold a standard in place.

Reach for it when the diagnosed cause looks like disorder rather than capability or capacity. If the data says operators cannot find what they need, or that two people run the same step two different ways, 5S gives you a structured way to design the fix. If the cause is demand variation or a supplier defect, a different Improve tool will serve you better.

The tool ships with a worked example, and this guide is built entirely around it. The case is fictional. It is a sample data set used to show how the tool behaves, not a report of real work.

The setting is an accounts payable area moving from paper capture to digital capture. The diagnosed cause is a workspace built for paper. Unused forms sit in cabinets. Capture stations are scattered. Clean-up happens when someone remembers. Methods are ad-hoc. Nothing audits the standard.

That is five separate symptoms of one cause. 5S lets you address each of them without treating them as five unrelated projects. Each symptom maps to a pillar, and each pillar gets one action. The sample uses five rows, one per pillar, which is the cleanest way to start.

The worksheet has six columns. Pillar, current state, future-state action, owner and target date. The discipline is in the wording. Current state should be observable. Future-state action should be a thing someone does, not an intention.

In the sample, Sort records a current state of unused forms and cabinets, with the action Remove and digitise, owned by the AP Lead, targeted 2025-04-08. Set in Order records scattered capture stations, with a centralised digital station, owned by Facilities, targeted 2025-04-10. Shine records inconsistent clean-up, with a daily digital workspace reset, owned by the Team, targeted 2025-04-12.

Standardise records ad-hoc methods, with standard SOPs for digital flow, owned by the Process Analyst, targeted 2025-04-15. Sustain records no audits, with a monthly 5S audit cadence, owned by the MBB, targeted 2025-04-30. The dates run in pillar order, which is sensible. You cannot standardise a layout you have not built, and you cannot audit a standard that does not exist yet.

The tool builds a delivery and ownership view from the worksheet. It takes the target dates and checks them against a reference date. In the sample that reference date is 2026-10-04, well after every target in the plan.

The headline is blunt. Five of five activities are past the target date, and the worst is 544 days past, which is Remove and digitise. Past target date reads 5. No owner reads 0. Complete reads 0 of 5. Status is Not set for all five rows.

Each row then repeats the detail. Remove and digitise, 2025-04-08, AP Lead, 544 days past. Centralised digital station, 2025-04-10, Facilities, 542 days past. Daily digital workspace reset, 2025-04-12, Team, 540 days past. Standard SOPs for digital flow, 2025-04-15, Process Analyst, 537 days past. Monthly 5S audit cadence, 2025-04-30, MBB, 522 days past.

Two signals matter here. Ownership is clean, because no row is missing an owner. Delivery is not, because nothing is marked complete and every date has passed. The tool also offers an AI review that explains the result in plain terms if you want a second read before you take it to a sponsor.

5S does not hand you a dollar figure. What it hands you is a scoped change with five actions, five owners and five dates. That is enough to size the gain by hand before you commit to a pilot.

Work row by row. Ask what Remove and digitise is worth if it lands, and what it costs in hours. Ask the same of the centralised digital station. The actions are small and separable, so each one can be estimated on its own. If the first two rows carry most of the value, you can pilot those two and hold the rest.

The delivery view then tells you whether the plan is live or dormant. In the sample, zero of five are complete and the worst row is 544 days past the target date. That is not a reason to abandon the design. It is a reason to re-date the plan and confirm the five owners still hold their rows before you count any gain.

The first decision is pilot or park. If the actions are clear and the owners are named, you have a change worth piloting in one area before you roll it wider. The sample passes the ownership test. It fails the delivery test, so the honest call is to re-plan, not to scale.

The second decision is sequencing. Five actions rarely start at once. The target dates in the sample give you an order. Sort first, then Set in Order, then Shine, then Standardise, then Sustain. If you need to compress, compress the middle, not the start.

The third decision is governance. The Sustain row exists so the gain does not evaporate. A monthly 5S audit cadence owned by the MBB is the control hook that Improve hands to Control and Sustain. Without it, the other four rows decay and you repeat the project in a year.

The first mistake is running 5S as a clean-up. Sort and Shine are visible, so they absorb all the attention. Standardise and Sustain are the pillars that hold the change. In the sample, those two rows carry the SOPs and the audit cadence, and they sit last for a reason.

The second mistake is vague ownership. The result tracks a no owner count for exactly this reason. The sample shows zero, which is what you want. Writing Team as the owner is acceptable for a daily reset that everyone performs, but it is weak for a design task.

The third mistake is leaving dates to drift and status to Not set. The sample shows all five rows at Not set and zero of five complete. The tool cannot tell whether the work stalled or whether nobody updated the sheet. That ambiguity is yours to remove.

The fourth mistake is applying 5S to a cause it cannot touch. If Analyse pointed at a capacity constraint, a centralised digital station will not fix it.

5S Workplace Design will not tell you whether the change is worth making. It holds no baseline, no measure of defects, no cost data and no cycle time. It will not calculate a saving, a yield or a payback. Any number you present has to come from your Measure work and your own estimates.

It will not verify that the actions were done. The complete count reflects what has been recorded in the worksheet, not what exists on the floor. In the sample, zero of five complete and five of five past the target date may mean the work stalled, or may mean nobody updated the status. The tool cannot separate those two cases, and neither can the AI review.

It will not choose the right pillar for a problem, and it will not stop you writing an action that does not match the current state beside it. It also assumes one row per pillar is enough. Real workspaces often need several rows under Set in Order alone.

Use it for what it does well. It converts a diagnosed cause into a designed future state with named owners and dated actions, and it shows you at a glance whether that plan is moving. Size the gain yourself, pilot in one area, then let the Sustain row carry the result into Control.