Line Troubleshooting

Why Your Conveyor Line Numbers Are Wrong: Common Mistakes and Fixes

The specific errors that make conveyor speed, throughput, and OEE numbers lie, with the symptom, root cause, and a numeric fix for each.

Symptom: your calculated conveyor speed says the line should hit 60 parts per minute, but the floor sees 42. Root cause is almost always mixing pitch and part length. If parts sit on 500 mm centers (pitch) but you fed 300 mm part length into a Conveyor Speed calc, your belt speed is off by 40 percent. The fix: belt speed equals pitch times rate, so 0.5 m times 60 per minute equals 30 m/min, not 18. Always drive speed off center-to-center spacing, never part size, and confirm the gap between parts is real, not assumed.

Symptom: takt says you have time to spare, yet you miss the shift target. The classic root cause is forgetting to subtract real breaks, changeovers, and planned stops before computing available time. A 480 minute shift is rarely 480 minutes. Strip two 15 minute breaks, a 30 minute lunch, and 20 minutes of startup and you have 400 minutes. Run Takt Time on 480 and you get 24 seconds per unit for 1,200 units; run it on the honest 400 and takt drops to 20 seconds. That 4 second gap is the difference between hitting and chasing the plan.

Symptom: OEE reads 92 percent but you are still short on good units. Root cause is double-counting or hiding losses. Availability, performance, and quality must each stay honest and separate. A common error is folding micro-stops under 2 minutes into performance instead of availability, or excluding rework from the quality factor. If true availability is 88 percent, performance 90 percent, and quality 96 percent, OEE is 0.88 times 0.90 times 0.96, which is 76 percent, not 92. Feed the OEE calc unfiltered stop logs, not the sanitized version the line lead reports.

Symptom: you speed up the conveyor and throughput barely moves. Root cause is chasing the wrong station. The line is paced by its slowest step, so accelerating the belt past the bottleneck just piles up work in process. If station 4 runs a 38 second cycle and everything else runs 25 seconds, use Bottleneck Capacity to find the true ceiling: 3,600 divided by 38 equals about 94 units per hour, period. Fix station 4 first. Line Balance analysis will show the 13 second imbalance you are paying for at every other station.

Symptom: your dwell time or cure calculation comes up short and product exits half-processed. Root cause is confusing loaded length with total length, or ignoring the accumulation zone. Dwell equals process length divided by belt speed. If the oven is 12 m but 2 m sits outside the heated zone, effective length is 10 m; at 5 m/min that is 2.0 minutes of real dwell, not 2.4. Run Dwell Time on the heated span only. A 20 percent overstatement here bakes in scrap you will not catch until QC rejects a full batch.

Symptom: machine utilization looks high but the line still starves. Root cause is confusing utilization with productive uptime. A machine can run 95 percent of clock time and still produce idling, blocked, or waiting on upstream feed. Machine Utilization measures busy time against available time; it says nothing about whether that busy time made good parts. Pair it with Cycle Time and Throughput so you separate a machine that is genuinely producing from one that is merely powered on and cycling empty carriers at 95 percent.

Symptom: schedule attainment swings 20 points week to week with no process change. Root cause is measuring attainment against a moving or padded target. If planners inflate the daily quota by 8 percent as a buffer, Schedule Attainment will read low even on a good day, and the number becomes noise. Lock the target to demonstrated Bottleneck Capacity, then measure actual good units against it. Attainment above 95 percent with a padded plan often hides a real number closer to 85, and you will make capital decisions on the wrong figure.

Symptom: two calculators disagree about the same line by 15 percent. Root cause is inconsistent time bases and units across tools. One calc runs on scheduled hours, another on calendar hours; one uses parts per minute, another per hour; feet and meters get swapped without conversion. Standardize before you compute: pick one time base, one length unit, and one rate unit for the whole line. A 3.28 factor between feet and meters, or a 60 factor between per-minute and per-hour, quietly turns a correct formula into a wrong answer every single time.

Published 2026-07-01.