Yamazumi: movable elements and constraints

Yamazumi: movable elements and constraints

Build each bar from observable work elements so a proposed change remains traceable. Here, operator work content is 85, 65 and 60 seconds against an 80-second takt. Moving E4, the ten-second empty-tote return, from operator one to operator two produces 75, 75 and 60 seconds. Total work remains 210 seconds. This is a theoretical redistribution, not a demonstrated new cycle. Check precedence, walking, skills, ergonomics, variability and machine overlap. The elements and their labels must stay visible in both views. Balanced bars can guide a trial; they do not by themselves justify a staffing decision.

Follow the method

1. Before
2. Move E4
3. Proposed
4. Validate

Reconcile E4 ownership before approving the trial

Fictional window-preparation teaching case: three trained operators share ten named elements. The team has 400 net available minutes and demand for 300 units. The supplied element times describe human work content for one unit, not observations proving sustained throughput.

Follow the method

1. 1
2. 2
3. 3
4. All

Move an element, then test the real work

Build and interpret work-element stacks, propose a feasible ownership change, and distinguish a balanced work-content chart from verified cycle performance.

Fictional window-preparation teaching case: three trained operators share ten named elements. The team has 400 net available minutes and demand for 300 units. The supplied element times describe human work content for one unit, not observations proving sustained throughput.

Role: Team leader facilitating a trial with the three operators and process engineer.

Normal condition

A proposed sequence must meet demand while retaining quality checks, safe access and the required precedence.

The gap

Operator 1 has 85 seconds of assigned work against 80-second takt; operators 2 and 3 have 65 and 60 seconds.

• Only E4, return empty tote, is provisionally transferable from operator 1 to 2.
• Walking, machine overlap, competence, access and variation need confirmation before adopting the proposal.

1. Establish the comparison

   Convert 400 minutes to 24,000 seconds and divide by 300 demanded units. Label the reference line 80 sec/unit and the vertical stacks human work content in seconds.

   Why: Using gross shift time or comparing minutes with seconds would create a false gap. Work content and elapsed cycle time answer different questions.

   Evidence: 24,000 /300 =80 seconds per required unit.

2. Keep each element identifiable

   Stack E1–E4 for operator 1, E5–E7 for operator 2 and E8–E10 for operator 3. Preserve element heights proportional to time and retain the E1–E10 legend.

   Why: A total bar shows overload but does not show which real piece of work could move. Removing a quality check just to lower a bar changes the method.

   Evidence: 85 +65 +60 =210 seconds of total human work.

3. Test the proposed allocation on paper

   Remove only the 10-second E4 segment from operator 1 and add it to operator 2. Keep the other nine elements, their durations and identities unchanged.

   Why: Redistribution changes ownership; it does not eliminate work. The total is a useful reconciliation against accidental omission.

   Evidence: 75 /75 /60 seconds; total 210. Nominal margins below takt are 5 /5 /20 seconds.

4. Check the physical dependency

   Walk the tote-return route with the operators under the local safe trial procedure. Confirm when the empty tote becomes available and whether operator 2 can reach it without delaying E5–E7 or crossing another task.

   Why: The arithmetic assumes E4 is transferable. Precedence and access can invalidate a visually attractive balance.

   Evidence: A trial plan names route, handoff point, permitted sequence, observer and stop conditions; actual results remain unfilled.

5. Compare prediction with observation

   Record multiple representative cycles and the element causing any wait, including mix, abnormal events and quality outcome. Compare observed performance with the 75 / 75 / 60 prediction before revising the standard.

   Why: An average stack below takt does not prove that variability or machine interaction permits every required cycle.

   Evidence: Retain timing sheets and a decision to adopt, adapt or stop; do not label the proposed chart an achieved improvement.

The route adds eight seconds

A hypothetical trial reveals that operator 2 needs 8 additional seconds of walking per unit to perform the tote return. This was absent from the supplied element model.

Update operator 2 to 83 seconds and total work to 218 seconds for this trial condition. Do not approve the original transfer unchanged; redesign access or test another authorized method.

83 exceeds 80-second takt by 3 seconds. Averaging the three operators together would conceal the local overload.

Record the extra walking as a new observed element, its timing basis and operating condition; retain the original proposal for comparison.

Run a practice session

Materials

• Ten E1–E10 element strips proportional to supplied times
• 80-second and 75-second takt references
• Blank allocation and trial record; calculator

1. Define units and demand · 5 minutes

   What does this vertical axis measure?

2. Demonstrate E4 transfer · 8 minutes

   Which work disappeared? None: reconcile 210.

3. Individual calculations, then pair challenge · 10 minutes

   Does zero nominal margin prove the new demand is feasible?

4. Debrief constraints and trial · 7 minutes

   What observation could invalidate your neat bars?

Debrief

• If totals change in the paper proposal, inspect element identity before recalculating.
• Separate a planning hypothesis from a measured standard and a staffing decision.

Draw stacks on a shared scale, complete the changed-demand problem, then compare reasoning with the answer.

Transfer into the work

Owner: Team leader and affected operators, supported by process engineering

Record: Element observation sheet, proposed assignment and trial decision

Review: Before changing the standard and after representative trial cycles

Evidence: Element timing/variation, safe access, precedence, quality and actual cycle behavior

Retain the authorized method, address the limiting condition and revise the trial; escalate systemic capacity gaps through daily management.