Cell Staffing and Human–Machine Balance

Lesson 16.4
12 minOpsida Teknoloji · Last reviewed: October 30, 2026

In one sentence

Human work gives the lower bound of the number of workers; the real arrangement is verified with each person's load and each machine's timeline.

By the end of this lesson

  • Calculate that 180 s/unit of human work needs at least 3 people for a 60 s/unit takt time.
  • Show human work, automatic processing and mandatory waiting in separate resource scopes.
  • Calculate the theoretical output interval from the longest load in the worker distribution.
  • Identify an arrangement where indivisible tasks or machine occupation make the lower bound insufficient.

Why it matters

To find the number of workers in a cell, first add up the work the person really does. Loading, inspection, walking and regular support work can be in this scope. The time a machine runs on its own is not added directly to the human load if the person can do other work at the same time. If, however, the person has to stay at the machine because of a safety or process condition, this attachment is shown separately. Separating human work from the time spent on the resource prevents showing the worker need as lower or higher than it is. The calculation belongs to a defined product and work sequence; automatic time that is free for one product may not be usable in the same way for another. First the right scope, then the division.

Dividing the total human work by the takt time produces a theoretical lower bound. If the result is fractional, the number of workers is rounded up. But a whole-number result does not prove a feasible distribution either. Tasks may be indivisible, two jobs may want the same person at the same time, or the machine may have to finish before the next job can start. Some walks arise anew when the worker distribution changes. For this reason the lower bound is used to test a target; it is not a firm staffing plan. If the number obtained cannot be reached, the arithmetic does not have to be wrong. The structure of the tasks and their precedence relations reveal conditions that the total load hides.

In human–machine balance, the aim is not to show all resources full at every moment. While the machine works automatically, it can be useful for the person to move to another safe job. It is also acceptable for a machine to stand idle for a while if demand is met. Filling every gap by adding work can force the worker to be in two places at once. The utilization percentage of resources must be read together with the output interval. If the longest person load exceeds the takt time, the flow does not keep up even if the total human load looks suitable. If the machine occupation exceeds the takt time, balancing the people perfectly does not close the gap. These two limits are calculated without confusing them and are combined on a shared time axis.

The staffing decision covers the safe and sustainable application of the work. Moving tasks between rows is not enough without defined training, skill authorization and material presentation. If a time with an allowance is used, adding the same allowance a second time is avoided; if a raw measurement is used, the required time scope is stated clearly. An exactly equal distribution at takt time does not by itself provide a margin for variation or additional regular work. In this lesson the fixed-time calculation is set up first, then its limits are shown. In the end the student can explain with numbers the difference between saying that three people are needed and proving that three people can really perform a specific work sequence.

Formula

Theoretical lower bound of workers (people) = ⌈Total human work (s/unit) ÷ Takt time (s/unit)⌉

Person load (s/unit) = Sum of assigned human jobs

Lower bound of human distribution (s/unit) = Longest person load

Distribution efficiency (%) = Total human work ÷ (Number of workers × Verified output interval) × 100

Machine occupation (s/unit) = Loading + Automatic processing + Other jobs holding the resource

Loads of 70, 61, 59 and 66 s at four stations; cycle time 70 s, takt time 72 s.70St. 1Gap 061St. 2Gap 959St. 3Gap 1166St. 4Gap 4Takt 72 sCycle 70 sTotal work: 256 s · 4 × 70 = 280 sEfficiency: 91.43%Balance loss: 8.57%
In this separate human distribution, the total is 256 s, the cycle time is 70 s and the balancing efficiency is 256/(4×70)×100=91.43%; the balance loss is 8.57% and the gaps are 0, 9, 11 and 4 s. Machine timelines are verified separately. This dataset was generated synthetically for training; it does not represent a real business, employee or customer study.

Unlock the rest of this lesson with an Academy account

Create a free Academy account with your work email address; it can be used once it is approved.

  • Step-by-step instructions
  • Worked example with real numbers
  • Where the method misleads
  • Common mistakes
Create a free Academy account