A yamazumi chart is a stacked bar chart that shows each operator's total work content as blocks piled up against the takt time line, so you can see at a glance who is overloaded, who has slack, and how to redistribute the work. Yamazumi is Japanese for "to stack up," and the chart is the standard tool for balancing an operation and freeing whole positions.

The power of the chart is that it makes an invisible problem visual. "Station three is always behind" becomes a bar taller than the takt line, sitting next to three bars that fall well short of it. This guide covers how to read a yamazumi chart, how to build one, and a worked rebalance that frees an operator. It is the visual companion to line balancing and depends on the same pace-setter, takt time.

What is a yamazumi chart?

A yamazumi chart is a vertical bar chart with one bar per operator or station, where each bar is built from stacked blocks representing individual work elements and their times. A horizontal line across the chart marks takt time, the pace the operation must hit to meet demand. The whole picture answers one question fast: is each person's work above or below the pace, and by how much?

Because the blocks are individual work elements, the chart does more than show who is overloaded. It shows exactly which chunks of work could move, and where they could go, because a block short enough to fit in another operator's headroom is a candidate to relocate. That is what turns a yamazumi chart from a diagnosis into a plan: it names the specific work you can shift, not just the fact that a shift is needed.

Yamazumi chart before rebalancingBefore: work stacked against takt, badly balancedTAKT 47sOP 1 - 40sOP 2 - 58sover taktOP 3 - 28sOP 4 - 25sEach block is a work element; OP2 is over takt while OP3 and OP4 sit half-empty
The unbalanced starting point: one operator over takt and two barely half-loaded. The gaps below the takt line are the headroom you have to work with.

How do you read a yamazumi chart?

You read a yamazumi chart by comparing every bar to the takt line and looking at the gaps. A bar above the line is an operator who cannot keep up, and the operation runs at their pace; a bar well below the line is paid capacity sitting idle. The distance between the tallest bar and the takt line is the immediate problem; the distance between the shorter bars and the line is the opportunity to fix it.

Color adds a second layer of insight. Many teams shade each work element by type: value-adding work in one color, necessary-but-non-value work (walking, inspecting, waiting) in another. A bar that is tall mostly because of non-value blocks is telling you to eliminate work, not just move it. This is where the yamazumi chart connects to the wider fight against muda, mura, and muri: the goal is not only a level stack but a shorter one, with the waste blocks removed before the value blocks are redistributed.

Color-coding a yamazumi bar by work typeColor tells you whether to move work or delete itTAKTVALUEwalk / inspectwasteremove firstVALUEtrimmednow under taktbefore
Shading each block by type shows whether a tall bar needs work removed (waste, excess walking) or simply redistributed. Cut the waste blocks before moving the value blocks.

How do you build a yamazumi chart?

You build a yamazumi chart by breaking the work into timed elements, stacking them per operator, and drawing the takt line across the top. It is deliberately low-tech; many of the best yamazumi charts are sticky notes on a whiteboard, one note per work element, so the team can physically move blocks during the rebalance.

  1. Calculate takt time. Net available time divided by demand sets the line every bar is measured against, per the takt time method. Without it, the chart has no reference and shows nothing actionable.
  2. Break each operator's job into work elements. Small, coherent chunks of work, typically a few seconds to under a minute each, including walking, reaching, and inspection, not just the value-adding steps.
  3. Time every element with real observation. Watch several cycles and, ideally, several operators, or pull actual cycle times from station data. Standards written years ago are the classic source of a chart that lies.
  4. Stack the elements into a bar per operator. One block per element, sized to its time, piled up. Color-code by value-adding versus non-value if you can; it changes the conclusions.
  5. Draw the takt line across all bars. Now the overloaded and underloaded operators are obvious, and the headroom under each short bar is visible and measurable.
  6. Move blocks to level the load. Relocate work elements from the tallest bars into the headroom of the shorter ones, respecting which tasks can legally move, until every bar sits just under takt. Eliminate the non-value blocks first.
  7. Lock the new balance as standard work. Rewrite each operator's standard work to the new element assignments, then run it, measure it, and adjust what the floor reveals.

What does a worked rebalance look like?

A worked example makes the method concrete. Take the four-operator operation in the chart above (illustrative numbers, not data from a real plant). Takt is 47 seconds. The starting loads are OP1 at 40s, OP2 at 58s, OP3 at 28s, and OP4 at 25s, totaling 151 seconds of work content. OP2 is over takt, so the operation cannot meet demand; OP3 and OP4 are barely half-loaded.

The theoretical minimum operator count is total work content divided by takt: 151 ÷ 47 = 3.2, rounded up to 4 as it stands, but the slack says otherwise. Move OP2's overflow down to OP3, then consolidate OP3 and OP4's light loads. Redistribute to three operators at roughly 50s each and you are over takt; but eliminate the non-value walking blocks first (say 10 seconds of avoidable motion across the operation), and 141 seconds of remaining work fits three operators at 47s. The result: three operators loaded just under takt, and a fourth freed for other work.

Yamazumi chart after rebalancingAfter: level bars under takt, one operator freedTAKT 47sOP 1 - 46sOP 2 - 47sOP 3 - 45sFREEDOP 4 - moved141s of work over 3 operators at ~46s each, after cutting 10s of non-value motion
After the rebalance: three level bars just under takt and a fourth operator freed, achieved by moving blocks and eliminating non-value motion, not by working faster.
OperatorBefore (sec)After (sec)Status
OP 14046Under takt
OP 258 (over)47At takt
OP 32845Under takt
OP 4250Freed for other work
Total15114110s non-value cut
The rebalance meets a 47-second takt with three operators instead of four, and trims 10 seconds of non-value work in the process.

By the numbers. The yamazumi (operator balance) chart is a documented tool in the Toyota Production System lineage catalogued by the Lean Enterprise Institute, used to allocate work elements evenly against takt (Lean Enterprise Institute, Operator Balance Chart). The stakes are labor: with U.S. manufacturing holding hundreds of thousands of open production jobs (BLS, Manufacturing: NAICS 31-33), freeing a fully-loaded operator through balancing is capacity reclaimed without a hire.

How is a yamazumi chart different from line balancing?

A yamazumi chart is the picture; line balancing is the practice. The chart is how you see the imbalance and plan the moves; line balancing is the broader discipline of doing it, including the precedence rules that decide which blocks can legally move. In practice you almost always use them together: the yamazumi chart is where a balancing exercise happens, block by block, in front of the team.

It also differs from a value stream map. A value stream map looks at the whole flow across the plant and finds where waiting and inventory pile up between processes; a yamazumi chart zooms into a single operation and balances the work among its operators. Use the map to find the process worth fixing, then the yamazumi chart to fix it. The waiting the map reveals is often the same idle time the yamazumi chart shows as gaps below the takt line, which is why cutting waiting waste and balancing a yamazumi chart are frequently the same job.

What are the common yamazumi mistakes?

The recurring mistakes are all shortcuts. Balancing on standard times instead of real observed times produces a chart that balances on paper and stalls on the floor. Moving value blocks around while leaving the non-value blocks in place levels the load without shortening it, so you meet takt with more labor than you need. Ignoring precedence, which tasks must precede others, yields a plan the floor cannot physically run. And treating the rebalance as a one-time event lets the balance decay the moment demand shifts takt or a process change alters an element.

The deeper issue is stale data. A yamazumi chart is only as honest as the element times feeding it, and on most floors those times are guessed or years old. Plants with live station-level visibility can pull actual cycle times per station per shift, which turns the yamazumi chart from an annual whiteboard exercise into a living picture that updates as the floor changes. See how one plant made its station times visible in our CLS case study. Pair a current yamazumi chart with the balancing discipline and a well-designed cell and rebalancing stops being a project and becomes a habit. No rip-and-replace; the stack just tells the truth.