Mixed-model line balancing: calculations and example
In one sentence
In mixed-model line balancing, the demand-weighted average explains total workload; each product's station workload and the time effect of consecutive demanding products are tested separately.
By the end of this lesson
- Calculate product-level and station-level weighted workloads for a 60% A and 40% B mix.
- Show the time deficit that three consecutive B products create under a takt time of 60 s/unit.
- Explain numerically that a change of sequence does not reduce work content.
- Verify a product-specific task transfer in terms of the average, the longest workload and task constraints.
Why it matters
Mixed-model line balancing considers both the shared and the different work of products that share the same line. Simply adding product times arithmetically is not enough; their shares in the order must be used. If 60% of demand is A and 40% is B, the average is 0.60 times A's time plus 0.40 times B's time. A simple average gives the right demand weighting only when quantities are equal. This calculation helps in understanding the total work needed over a planning period. But no operator produces the average product; at any moment they face either A or B. The average workload calculation therefore does not replace a product-level time check. The two must be read together.
If the work at a station takes 50 s/unit for A and 75 s/unit for B, the average for the stated mix is 60 s/unit. With a takt time of 60 s/unit as well, the average fits the target exactly. Even so, product B exceeds its allotted time by 15 s/unit. When long products arrive back to back, this deficit accumulates. That the time short products would have provided was spent idle earlier does not retroactively complete a long job that comes later. Demand share, product sequence and the station's working time are therefore tracked in separate records. A daily total that looks suitable does not prove that individual products are completed on time. Examining demanding sequences in particular reveals the load that the average hides.
Smoothing the product sequence does not eliminate long jobs. Placing A between B products may limit how much the accumulated deficit grows; however, a resource to do the work and a feasible timing arrangement are still needed. If the line makes mandatory transfers at fixed moments, the assumption that a station can hold back its delay may not hold. If there is intermediate stock or a flexible completion window, their physical and safe limits are shown explicitly. Do not behave as if a buffer exists automatically just because the average is suitable. Sequence is a condition of balancing, not a method change that shortens time. Replenishment, product changeover and tool setup are also part of the feasibility of this sequence, and the relevant resources must be visible in time.
A sound solution explains what the same task allocation means for each product. If a B-specific task is to be moved to the adjacent station, check whether it conflicts with the tasks of product A. If the task does not exist on that product, zero can be written in the workload cell; an unmeasured time cannot be given zero. When the product mix changes, the averages are recalculated. The most difficult product or sequence is also retested; old order shares are not carried over to new demand. In this lesson, the acceptance decision rests on product workloads and demanding-sequence conditions, not only on the period average. The aim is to explain how the same line handles different products within real task and time limits.
Formula
Product share = Product demand (units/period) ÷ Total demand (units/period)
Weighted workload (s/unit) = Σ(Product share × Product workload)
Lower bound on operator count (people) = Average total work ÷ Takt time, rounded up to a whole number
Product workload gap (s/unit) = Product workload − Takt time
Accumulated time deficit (s) = max(0, Previous deficit + Next product workload − Takt time)
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