Line Balance

Line Balancing as a Weekly Operating Practice

A line is only as fast as its slowest station, and everyone else is paid to wait. This playbook covers balance efficiency math, rebalance levers, and the weekly practice that keeps idle seconds from calcifying.

An unbalanced line pays people to wait, every cycle, all shift. If five stations carry 50, 45, 60, 40, and 55 seconds of work, the line moves at the 60 second bottleneck and the other four stations idle a combined 50 seconds per cycle. Over a 450 minute shift, that is 450 units and 6.25 labor hours of paid waiting per day, roughly 200 dollars daily or 50,000 dollars a year at 32 dollars fully loaded, on one small line. Balance losses are invisible because everyone looks busy; the math is the only witness.

Measure balance with three numbers. Balance efficiency equals total work content divided by station count times bottleneck time: 250 seconds over 5 times 60, or 83.3 percent. Balance delay is the complement, 16.7 percent, which is pure idle. Bottleneck time, 60 seconds here, sets line output: 27,000 net seconds gives 450 units, and no rebalance changes that unless work leaves the bottleneck station. The Line Balance calculator computes efficiency, bottleneck, and idle time per station from a simple list of station times, so the study reduces to timing 25 to 30 cycles per station honestly.

Benchmark ranges are well established. Assembly lines in good shape run 85 to 95 percent balance efficiency; below 80 percent means at least one station's worth of labor is being paid to idle. Above 95 percent is suspect unless variation is very low, because zero slack turns every hiccup into a line stop; at 10 percent cycle variation, target 88 to 92 instead. Also watch the spread: a bottleneck more than 15 percent above mean station time flags a station that was never properly decomposed, and idle concentrated in one station beats idle scattered thinly, because concentrated idle can be redeployed.

The rebalance levers, in order of cost: move work elements across station boundaries, which is free where precedence allows; shifting a 12 second element off the 60 second bottleneck to the 40 second station takes the line to a 55 second pace and adds 9 percent output. Re-sequence elements to break false precedence constraints, which commonly hide 20 to 30 percent of feasible moves. Offload machine-tended time so one operator covers two stations. Add duplicate capacity at the bottleneck only as the last resort. And balance to takt, not to the fastest achievable bottleneck, whenever demand pace is the binding number.

Failure modes repeat everywhere. Balancing once at launch and never again, while engineering changes add 3 to 5 seconds a year to random stations until efficiency quietly falls from 90 to 78 percent. Balancing to average times when station variation differs wildly, so the paper balance jams in practice. Timing the best operator, then staffing with the crew you actually have. Ignoring precedence and rework loops, producing balances that cannot be built. And the worst: treating the balance chart as an industrial engineering artifact instead of a supervisor's daily tool, so nobody on the floor knows which station owns the pace.

Make balancing a weekly practice, not an annual project. Daily, the hour board at the bottleneck station tells you if the balance is holding; misses above 5 percent get coded. Weekly, re-time the bottleneck and any station flagged by operators, update the yamazumi chart, and execute one work element move if the math supports it; small weekly moves beat quarterly overhauls because operators absorb single-element changes without a learning curve dip. Monthly, run a full re-time of all stations. On any takt change above 10 percent or any engineering change touching work content, rebalance before the first shift runs the new state.

World-class balance practice shows up as numbers and habits: balance efficiency held above 90 percent across takt changes, rebalances executed within 5 working days of a demand shift, yamazumi charts posted at the line and current within a week, and team leads who can move two work elements themselves without an engineer. Plants at this level typically staff 8 to 12 percent leaner at equal output and flex 20 percent up or down in volume within a week. The line stops being a fixed machine and becomes an instrument the team retunes as demand moves.

Published 2026-07-02.