Common Mistakes

Costly Mistakes That Wreck Elevator and Escalator Manufacturing Numbers

A troubleshooting guide to the recurring errors that throw off cab takt, rail take-offs, motor margins, and test tower plans, with the symptom, root cause, and numeric fix for each.

The single most expensive mistake in cab line planning is feeding gross shift length into takt instead of net available time. Symptom: the line paces to a 480 second takt but WIP still stacks between car-sling and dress-out. Root cause: 30 minutes of breaks, startup, and meetings never get subtracted, so the true takt is nearer 450 seconds. Fix: strip non-productive minutes before running the Elevator Cab Assembly Takt calculator. On a 60 cab shift that 30 minute error inflates apparent capacity by about 6.7 percent, roughly four cabs of phantom throughput you will never build.

Ordering guide rail short is the classic material take-off failure. Symptom: the crew runs out of T-rail at the top splice and the hoistway stalls waiting on a reorder. Root cause: estimators use floor-to-floor height, not full pit-to-overhead travel, and count only the two car rails while forgetting the two counterweight runs. On a 165 foot job that drops the count from four runs to two and undershoots by 660 feet before waste. Fix: run Guide Rail Length Planning with all four runs and a 1.05 to 1.10 waste multiplier, then round up to whole 20 foot sticks.

Motor sizing goes wrong when engineers read a positive margin as proof the machine is safe. Symptom: a drive that shows healthy headroom on paper still overheats and trips under duty. Root cause: the margin figure is a steady-state capacity ratio and ignores acceleration torque peaks and duty-cycle heating. A machine at 42 kW available against 35 kW demand looks fine at 20 percent margin, but repeated 90 percent-load starts can push instantaneous demand well past rating. Fix: treat the Elevator Motor Load Margin result as a first screen only, then verify thermal and dynamic load separately before releasing the drive package.

Double-counting door events silently doubles endurance numbers. Symptom: a test log reports 36,000 accepted cycles when the operator only saw 18,000 full open-close events. Root cause: the protocol logs one open plus one close as two events, but the spec already defines a full open-close as a single cycle, so the events-per-cycle input is set to 2 when it should be 1. Fix: reconcile the counting basis before running Elevator Door Cycle Count. Getting this wrong overstates validated life by 100 percent and can ship an operator with half its proven endurance.

Committing ship dates against raw slot count is where test planning breaks. Symptom: sales promises 36 units this month and the plant delivers 30. Root cause: the schedule ignores tower uptime and first-pass yield. At 88 percent uptime and 96 percent first-pass yield, 36 gross slots yield only about 30.4 accepted units, with 4.3 lost to downtime and 1.3 held for retest. Fix: quote against the Test Tower Capacity accepted figure, not the slot grid. Every retested unit re-occupies a slot, so low yield compounds the shortfall rather than adding to it linearly.

Field labor bids leak margin when scope capture is left at 100 percent by default. Symptom: the install total looks competitive but the crew is only setting car and rails while a GC handles hoistway construction. Root cause: the estimate prices hours the contractor never performs. On 420 crew hours at a 95 dollar loaded rate, dropping capture from 100 to 70 percent changes variable labor from 39,900 to 27,930 dollars, a 12,000 dollar swing. Fix: set the real scope share in Field Install Labor Cost and confirm mobilization and supervision are carried as a separate fixed line, not buried in the hourly rate.

Custom options get under-recovered because estimators price only material and forget engineering and slower assembly. Symptom: a job wins on price then bleeds margin through the shop. Root cause: a non-standard finish or fixture adds detailing hours and line disruption that a per-unit material adder never captures. If eight options each carry 650 dollars of burden plus a 4,000 dollar fixed engineering charge, that is 9,200 dollars the base configuration never sees. Fix: quantify each adder with Custom Option Burden and confirm the total flows into Vertical Transport Unit Cost rather than getting absorbed as overhead.

Warranty exposure gets misjudged when teams average severity, occurrence, and detection instead of multiplying them. Symptom: a unit scores mid-range and ships, then generates callbacks that erode the job's margin. Root cause: a defect that is severe (8), likely to recur (6), and hard to catch pre-shipment (7) multiplies to 336 on a 1 to 1000 scale, but a naive average reads 7 and looks tame. Fix: use the Warranty Exposure Score weighting and gate release on the compound number. A single field callback on a shipped elevator commonly costs several crew-days plus parts, dwarfing the inspection time that would have caught it.

Published 2026-07-01.