Method guide · Value stream mapping and flow
How many operators does this line need? Takt, work content, balance and relief
Short answer
Divide the total work content of one unit by takt time and round up. For example, 252 seconds of work at a 60-second takt is 4.2, so at least 5 operators. Then check two things: that the work splits into stations in sequence with the longest one under the time you have per unit, and that the losses you know about, such as stoppages, are in the sum as a factor you set from your own records. The roster is larger again: divide the operators per shift by the share of people present on an average day, round up for each crew, and make sure every station has trained cover.
The example line: 17 work elements, 252 seconds
Every number on this page comes from one line, so you can follow each sum. The line is illustrative: a manual line assembling a small electric water pump, five stations in a row, two shifts. Each time is the operator's lowest repeatable time over ten timed cycles, the lowest time that came up at least twice, as on the time study sheet. Only operator time counts: the leak tester runs for 45 seconds on its own, so element 14 is the operator's 14 seconds to load, start and unload it.
| # | Work element | Seconds | Type | Station before | Station after |
|---|---|---|---|---|---|
| 1 | Pick housing from flow rack, place in fixture | 10 | Necessary | 1 | 1 |
| 2 | Press in shaft seal | 16 | Value-added | 1 | 1 |
| 3 | Fit shaft and impeller | 18 | Value-added | 1 | 1 |
| 4 | Walk to the rack for cover screws | 8 | Waste | 2 | 1 |
| 5 | Fit O-ring and cover | 16 | Value-added | 2 | 2 |
| 6 | Drive six cover screws | 26 | Value-added | 2 | 2 |
| 7 | Re-check cover screws with a hand torque wrench | 8 | Waste | 2 | 2 |
| 8 | Unpack motor from its carton | 8 | Waste | 2 | 3 |
| 9 | Fit motor to housing | 14 | Value-added | 3 | 3 |
| 10 | Drive four motor bolts | 22 | Value-added | 3 | 3 |
| 11 | Fit cable gland | 6 | Value-added | 4 | 3 |
| 12 | Connect wiring | 26 | Value-added | 4 | 4 |
| 13 | Fit terminal box lid | 12 | Value-added | 4 | 4 |
| 14 | Load, start and unload leak tester (it runs 45 s on its own) | 14 | Necessary | 4 | 4 |
| 15 | Electrical safety test: connect, read, disconnect | 18 | Necessary | 5 | 5 |
| 16 | Label and final visual check | 12 | Necessary | 5 | 5 |
| 17 | Bag, box and place on outbound conveyor | 18 | Necessary | 5 | 5 |
| Total | Work content for one pump | 252 |
Types follow the yamazumi convention: value-added, necessary but not value-added, and waste. The times are illustrative, not from a real line.
Step 1: takt time from demand and available time
Takt time is the available production time divided by customer demand for the same period. Available time is the time the line is staffed and expected to run: the paid shift minus scheduled breaks and meetings, when the line stops for them.
| Item | Per shift | Per day, 2 shifts |
|---|---|---|
| Paid shift time | 480 min | 960 min |
| Less two 10-minute breaks (the line stops) | 20 min | 40 min |
| Less the start-of-shift meeting | 10 min | 20 min |
| Available time | 450 min (27,000 s) | 900 min (54,000 s) |
| Customer demand | 450 pumps | 900 pumps |
| Takt time | 60 s | 60 s |
Use the demand for the period you are staffing, such as next month's daily average, and work it out again when demand changes. The Lean Enterprise Institute's lexicon notes that Toyota reviews takt every month. The takt time calculator does this sum for you.
Step 2: the theoretical minimum is work content divided by takt
Total work content is the operator time for every element of one unit added up: 252 seconds here. Divide it by takt and round up, because you cannot staff 0.2 of a person. Groover calls the result the theoretical minimum number of workers: the smallest whole number at or above work content divided by cycle time.
The theoretical minimum for the example
Example numbersOperators = total work content ÷ takt time, rounded up
The smallest whole number at or above the result.
- Takt =2 × 450 min × 60 ÷ 900 pumps= 60 s
- Work content =sum of the 17 element times= 252 s
- Operators =252 ÷ 60= 4.2
It is a floor, not a plan. It assumes the work splits into equal slices, nobody waits for anybody, and the line never stops. Steps 3 and 4 test those assumptions, and they are where most staffing plans go wrong.
Step 3: put stoppages, allowances and relief in as factors you set
The minimum uses every second of available time, and real lines lose some of it. Put each loss into the sum as a number you choose and write down where it came from. That is better than adding a person to be safe, because the next person to look at the plan can see why the number is what it is.
| Factor | Where it goes | In the example | Where your number comes from |
|---|---|---|---|
| Scheduled breaks and meetings | Out of available time in step 1, when the line stops for them | 30 min a shift, already out of the 450 | The shift pattern |
| Personal needs and fatigue allowance | Onto the element times, only if scheduled breaks do not already cover it | None: the line stops for scheduled breaks | Your site's time study practice. Standard times from a time study already include the allowance |
| Stoppages: breakdowns, waiting for parts, quality holds | Off the time you have per unit: takt × (1 minus loss) | 5%, so 60 × 0.95 = 57 s per unit | Your downtime log over a normal month |
| Break relief, if the line runs through breaks | Extra people, not extra time (step 6) | Not needed: the line stops | The break schedule |
Count each loss once. A common double count is taking breaks out of available time and then adding a fatigue allowance for the same rest to the element times.
The same sum with a stoppage factor
Example numbersOperators = work content × (1 + allowance) ÷ (takt × (1 minus stoppage loss))
Allowance 0 here, because the line stops for its scheduled breaks. Stoppage loss 5%, an example value.
- Time per pump =60 × (1 minus 0.05)= 57 s
- Operators =252 × (1 + 0) ÷ 57= 4.42
With the 5% stoppage factor, each pump has 57 seconds, not 60, and the minimum becomes 252 ÷ 57 = 4.42, still 5 operators. In general: operators = work content × (1 + allowance) ÷ (takt × (1 minus stoppage loss)), rounded up. Freivalds' textbook works the same way with an expected efficiency: its example divides by 95% before rounding up.
Step 4: split the work into stations and check the longest one
Five operators is the number. Whether five works depends on how the 252 seconds split into stations, in the order the pump is built. The line can only run as fast as its longest station, so check that station against the 57 seconds per pump, not the average.
Two measures say how good the split is. Balance efficiency is the total work content divided by the number of operators times the longest station time. Balance delay, also called balance loss, is 1 minus balance efficiency: the share of paid operator time spent waiting for the slowest station. Groover and Freivalds both define it this way, with every station allowed the slowest station's time.
The example line as a yamazumi, before and after
Example numbersBefore: 5 operators
After: 5 operators
Option: 4 operators
- Value-added
- Necessary, not value-added
- Waste
- Takt 60 s
- 57 s per unit after stoppages
- Numbers in the blocks are element numbers; totals in red are over 57 s.
Three yamazumi charts. Before rebalancing, five stations of 44, 66, 36, 58 and 48 seconds, with station 2 over the 60-second takt. After rebalancing, five stations of 52, 50, 50, 52 and 48 seconds, all under 57 seconds. The four-operator option has stations of 44, 56, 54 and 56 seconds.
| Station | Before: elements | Before: seconds | After: elements | After: seconds |
|---|---|---|---|---|
| 1 | 1 to 3 | 44 | 1 to 4 | 52 |
| 2 | 4 to 8 | 66 | 5 to 7 | 50 |
| 3 | 9 to 10 | 36 | 8 to 11 | 50 |
| 4 | 11 to 14 | 58 | 12 to 14 | 52 |
| 5 | 15 to 17 | 48 | 15 to 17 | 48 |
| Measure | Before | After |
|---|---|---|
| Longest station | 66 s (station 2) | 52 s (stations 1 and 4) |
| Fits 57 s per pump? | No: 9 s over | Yes: 5 s to spare |
| Pumps a day at the longest station, after 5% stoppages | 777, short of 900 | 986 |
| Balance efficiency: 252 ÷ (5 × longest) | 76.4% | 96.9% |
| Balance delay | 23.6% | 3.1% |
| Line efficiency in the line balancing calculator, takt 60 s | 76.4% | 84.0% |
The rebalance moved three elements and no work was removed. The walk for cover screws (element 4) went to station 1, unpacking the motor (element 8) went to station 3 next to the motor fitting it feeds, and the cable gland (element 11) went from station 4 to station 3. Station 2 fell from 66 to 50 seconds.
The two efficiencies answer different questions. Balance efficiency (96.9%) says the work is spread evenly. The line balancing calculator measures against takt when the longest station is under it (84.0%), because a line paced to demand still waits out the gap between its longest station and takt. Enter 57 as the takt to see the result with the stoppage factor in.
Move an element only if the build order still works, the parts and tools are in reach at the new station, and the operators who will do it agree it can be done. Try it at the line for a few cycles before you change the standard work.
Step 5: when the answer is 4.2, decide what to do with the 0.2
Rounding 4.2 up to 5 is the safe answer, but it pays for waiting: 5 operators × 57 seconds is 285 seconds of paid time per pump for 252 seconds of work, so 33 seconds of every pump is spent waiting somewhere on the line. The Lean Lexicon describes the aim of an operator balance chart as fewer operators, each with work very nearly equal to, but slightly less than, takt. There are four ways to handle the fraction, and they can be combined.
| Option | What it means here | The numbers | Watch out for |
|---|---|---|---|
| Add the fifth operator | Run 5 operators on the rebalanced stations | Longest station 52 s; 33 s of waiting per pump across the line | The waiting becomes normal. Use it for improvement work, and take the fifth person off when the work content falls |
| Remove the 0.2 with kaizen | Remove the three waste elements: screws presented at the station, motors delivered unpacked in returnable trays, and the duplicate torque check dropped with quality's agreement | 252 minus 24 = 228 s: 3.8 operators at takt, 4.0 at 57 s. In build order, the best split into 4 stations still has a 62 s station | On paper 4 fits; in sequence it does not. And 4.0 at 57 s would load every operator to 100% of the time there is |
| Share part of a person | Do the kaizen and move boxing (element 17, 18 s) to the material handler who already serves the line | 210 s left on the line: 44, 56, 54 and 56 s for 4 operators, all under 57 s. Boxing needs 18 × 900 = 16,200 s a day, 135 minutes a shift, on the handler's route | Only works if the handler's route has that time, and the shared work sits at the start or end of the line |
| Run 4 operators for longer | Keep 4 operators and work overtime | Even with a perfect split, 4 operators need 252 ÷ 4 = 63 s per pump: 900 pumps take 56,700 s, 2,700 s (45 minutes) a day more than the 54,000 s available, before stoppages | Daily overtime is a staffing decision made by default, and it leaves no time to recover from a bad day |
Balance efficiency for the shared option: 210 ÷ (4 × 56) = 93.8%.
Which option is right depends less on the decimal than on whether the work splits into stations that fit, and only the element table shows that. In the example, removing 24 seconds of waste was not enough on its own. Moving one 18-second element off the line is what made 4 operators work.
Step 6: from operators per shift to people on the roster
Operators per shift is how many people the line needs at any moment. The roster also has to cover absence: holidays, sickness, training and anything else that takes trained people away. Divide by the share of people present on an average day, and round up for each crew, because someone on the day crew cannot cover nights.
People per crew
Example numbersPeople per crew = operators per shift ÷ (1 minus absence rate), rounded up
Absence rate 8%, an example value. Round for each crew, not for the day.
- People =5 ÷ (1 minus 0.08)= 5.43
| Item | Per crew | Both crews |
|---|---|---|
| Operators needed at the line | 5 | 10 |
| Absence rate (example value; use your own from the absenteeism tracker) | 8% | 8% |
| People needed: 5 ÷ (1 minus 0.08) | 5.43, so 6 | 12 |
| The same sum done for the whole day instead | 10 ÷ 0.92 = 10.87, so 11, which leaves one crew with no cover |
If the line runs through breaks instead of stopping, available time goes up and takt gets longer, but every break needs a relief operator trained on that station. Count the minutes to cover each shift: 5 operators × 20 minutes of breaks is 100 minutes of relief a shift, and decide who does it.
The absence rate covers an average day. On a bad day it will not, and that is where cross-training matters. For each station, count the people on each crew who are qualified to work it alone. A job cover matrix is built for this, and two qualified people per job is a common starting point. With 6 people for 5 stations, a crew where everyone is qualified on at least two stations has 12 qualifications for the 10 it needs. Plan the training in the training matrix template or the cross-training plan, and keep the absence rate in the absenteeism tracker.
Mistakes that make the number wrong
- Using the best-case time. A single fastest reading, or the time of the quickest operator, makes the line look as if it needs fewer people than it does. Use the lowest repeatable time from several cycles timed with more than one operator, and keep the spread. The time study sample size calculator tells you how many cycles to time.
- Ignoring variation. An average hides the cycles that stop the line. In the example, if wiring takes 34 seconds instead of 26 on a stiff cable, station 4 goes from 52 to 60 seconds, over the 57 seconds per pump. Find the elements that vary most and fix the cause, or keep them off the station with the least spare time.
- Designing for 100% utilisation. A plan that loads every operator to exactly the time there is has no room for a dropped part or a slow cycle, so each one becomes a lost unit. Load each operator to slightly less than takt and know how many seconds each station has spare.
- Counting machine time as operator work. Only the operator's manual and walking time is work content. When a machine runs on its own, check its cycle separately on a process capacity sheet, and the operator's sequence on a standard work combination table.
- Counting breaks twice. Take them out of available time or add them as an allowance, not both.
- Rounding the roster for the whole day. Round for each crew, and check cover station by station, not only in total.
- Working it out once. When demand changes, takt changes, and the sum has to be done again, roster included. The Lean Lexicon calls staffing that moves with volume labor linearity; Toyota calls it a flexible manpower line.
Where the formulas come from
The arithmetic on this page is standard industrial engineering. These are the sources it was checked against. The example line and its numbers are ours and illustrative.
- Takt time, operator balance chart, labor linearity and the standardized work combination table: Lean Enterprise Institute, Lean Lexicon, the entry for each term.
- Theoretical minimum number of workers, balance efficiency and balance delay: Mikell P. Groover, Automation, Production Systems, and Computer-Integrated Manufacturing, 3rd edition (Prentice Hall, 2008), chapter 15, Manual Assembly Lines.
- Line efficiency as total standard minutes over total allowed minutes, and dividing by an expected efficiency before rounding up: Andris Freivalds, Niebel's Methods, Standards, and Work Design, 12th edition (McGraw-Hill, 2009), chapter 2, line balancing.
- Standard time, relaxation allowances for personal needs and fatigue, and contingency allowances: International Labour Office, Introduction to Work Study, 4th revised edition, edited by George Kanawaty (ILO, 1992), chapter 23.
- The workbook on designing cells and lines this way: Mike Rother and Rick Harris, Creating Continuous Flow (Lean Enterprise Institute, 2001), especially the chapters What Is the Work and Distributing the Work.
Go deeper
This page is the short version. These guides cover each part in full.
- Yamazumi Chart: Workload Balancing for JIT Production LinesBuilding the yamazumi chart step by step and using it in a kaizen event.
- The Three Elements of Standard Work: Takt Time, Work Sequence, Standard WIPWhat changes in the work sequence and standard WIP when takt changes.
- Standard Work Tools: Combination Sheet, Standardized Work Chart, Job BreakdownThe forms that hold each operator's sequence, with manual, walking and machine time.
- Why One Employee Absence Can Shut Down Production LinesWhy thin cover stops lines, and how to set cross-training priorities.
All our articles on this subject are in the value stream mapping and flow topic.
Free templates and tools
- Yamazumi chart template (operator balance)A printable yamazumi grid with the takt line, a task table per operator, and an Excel version that works out each operator's total, line balance efficiency and minimum operators.
- Time study sheetTime each element over 10 cycles and work out the average, lowest repeatable and standard time. Excel does the sums.
- Standard work combination tableOne operator's manual, machine, walk and wait time on a time chart against takt, with machine cycles and a process capacity sheet.
- Process capacity sheet templateEach machine in a part's routing with manual, auto and tool change time, its capacity per shift, and the step that limits the cell against demand.
- Training matrix template with expiry datesPeople by required trainings with expiry dates, OK, due, expired and missing flags, compliance by person and area, cover per training and a renewal list.
- Cross-training plan templateWho is learning which task, from and to which level, by when, and how many are qualified on each task.
- Absenteeism tracker templateA monthly attendance grid with codes, absence rate by person and team, and an optional Bradford factor.
- Takt time calculatorThe pace a line must hit to meet customer demand, from shift time, breaks and daily demand.
- Line balancing calculatorLine efficiency, balance delay and the minimum number of stations from station times and takt.
- Time study sample size calculatorHow many cycles of an element to time, from your first readings, with the t formula, the ILO formula and the range method.
- Cycle time calculatorCycle time against takt, units per hour, capacity per shift and the gap to daily demand.
Doing this in LeanSuite
Line Balancing
Drop in a video of your line and get a Yamazumi chart, a takt comparison and rebalancing recommendations.
See Line BalancingJob Cover Matrix
Shows who is qualified to cover each job, built from skill levels and signed-off procedures, so you can cover an absence fast and spot single points of failure.
See Job Cover Matrix
Open this board in LeanSuite
The yamazumi operator balance board is one of the 30 ready-made boards in Lean Studio, LeanSuite's whiteboard. Try the interactive demo in your browser with sample data, or pilot LeanSuite on your own lines for 90 days, with the setup done by our team.
What Lean Studio isFAQ
Questions about how many operators a line needs
Rather see it on your own problem?
Bring the problem you are stuck on, and we will show you on a call how LeanSuite handles it.

