How do I measure rework cycles per position-hour?

Time a representative sample of repairs at the bench and divide 60 by the average minutes per repair. Include diagnosis, re-test and paperwork, since each ties up the position. Weight mixed defects by their share of rework volume.

What if failed repairs get a second attempt?

Then each failure comes back as new load. Multiply the rework rate by 1 plus the failure share (1.25 at 75% recovery yield) and rerun the check to size positions for both passes.

Is 95% rework loop utilization too high?

It keeps up on average, but repaired units wait much longer. In Kingman's formula queue time scales with utilization ÷ (1 − utilization), so moving from 80% to 95% multiplies the average wait by about five.

Do good units returned count toward line output?

Yes, once they pass re-inspection they rejoin good stock, so final good output is first-pass good units plus good units returned. First pass yield excludes them, which is why it reads lower. Units scrapped after rework are the real loss.

Conveyors calculator

Rework Loop Capacity Calculator

Check whether a rework loop can repair units as fast as inspection sends them, and how many positions it needs. You need line rate, rework rate, positions, repair cycle rate, uptime and recovery yield.

What this calculator does

Whether a rework loop can repair units as fast as the line sends them, and how many come back good.

Formula used

Units sent to rework = line rate × rework rate
Effective rework capacity = positions × cycles per position-hour × uptime
Utilization = units sent ÷ capacity; spare capacity = capacity − units sent
Positions needed = FLOOR(units sent ÷ (cycles per position-hour × uptime)) + 1, the fewest keeping utilization below 100%
Good units returned = MIN(units sent, capacity) × recovery yield; the rest are scrapped

Inputs explained

Line Rate Through Inspection: Units per hour through the inspection that routes defects to rework.
Rework Rate: Units sent to rework ÷ units inspected, from your quality log.
Active Rework Positions: Repair benches or cells working the loop in parallel.
Rework Cycles per Position-Hour: Repairs one position completes per working hour, from time study.
Rework Loop Uptime: Share of scheduled time staffed and able to repair, from shift logs.
Rework Recovery Yield: Reworked units that pass re-inspection ÷ units reworked, from the rework log.

How to use the result

Best suited to staffing a repair bench after a defect spike, sizing a rework cell for a new line.
Uses hourly averages: a defect burst can overload the loop for hours while the shift average stays below 100%. Mixed defects with very different repair times need a volume-weighted average repair time.

Current U.S. benchmarks

The U.S. has 21,668 machinery manufacturing establishments employing about 1,086,146 workers (Census County Business Patterns, 2023).

Common questions

How do I measure rework cycles per position-hour? Time a representative sample of repairs at the bench and divide 60 by the average minutes per repair. Include diagnosis, re-test and paperwork, since each ties up the position. Weight mixed defects by their share of rework volume.
What if failed repairs get a second attempt? Then each failure comes back as new load. Multiply the rework rate by 1 plus the failure share (1.25 at 75% recovery yield) and rerun the check to size positions for both passes.
Is 95% rework loop utilization too high? It keeps up on average, but repaired units wait much longer. In Kingman's formula queue time scales with utilization ÷ (1 − utilization), so moving from 80% to 95% multiplies the average wait by about five.
Do good units returned count toward line output? Yes, once they pass re-inspection they rejoin good stock, so final good output is first-pass good units plus good units returned. First pass yield excludes them, which is why it reads lower. Units scrapped after rework are the real loss.

Last reviewed 2026-10-01.