How the roll-up gate works
How the Backlog-to-Capacity Lead-Time Reality Check works
This tool grades each shop work center's quoted lead time against what its demonstrated throughput can actually cover, then rolls the worst read up into one shop-wide verdict. It reads 7 work centers and rolls them into one verdict — the worst, never the average — then a gate that can only make it worse.
The 7 work centers
01SAW3.0 wks quoted vs 2.0 wks computed (120 hrs backlog / 60 hrs/wk throughput) - intake 58 hrs/wk, steadyCOVERED
02CNC-MILL5.0 wks quoted vs 6.0 wks computed (540 hrs backlog / 90 hrs/wk throughput) - 90 hrs short of covering the backlog, intake 88 hrs/wk steadyCUTTING IT CLOSE
03WELDfix first4.0 wks quoted reads COVERED on the ratio alone (192 hrs cover vs 96 hrs backlog) - but intake is 70 hrs/wk against only 48 hrs/wk of demonstrated throughput, +22 hrs every weekDIVERGING QUEUE
04PAINT2.5 wks quoted vs 2.0 wks computed (100 hrs backlog / 50 hrs/wk throughput) - intake 45 hrs/wk, steadyCOVERED
05ASSY2.5 wks quoted vs 3.0 wks computed (210 hrs backlog / 70 hrs/wk throughput) - 35 hrs short of covering the backlog, intake 66 hrs/wk steadyCUTTING IT CLOSE
06INSPECT2.0 wks quoted vs 1.5 wks computed (45 hrs backlog / 30 hrs/wk throughput) - intake 28 hrs/wk, steadyCOVERED
07BRAKE2.5 wks quoted vs 2.0 wks computed (84 hrs backlog / 42 hrs/wk throughput) - intake 40 hrs/wk, steadyCOVERED
LEAD TIMES REAL
every work center's quoted lead is covered by demonstrated throughput
QUOTING TIGHT
the worst center's cover falls within 75% of its backlog but still short (CUTTING IT CLOSE)
LEAD-TIME FICTION
any center's cover falls below 75% of its backlog - or any center's weekly intake exceeds its demonstrated weekly throughput (a diverging queue)
Diverging-queue gate · worsen-only
If any work center's weekly intake exceeds its demonstrated weekly throughput, its backlog is growing every week. A lead time computed from a growing queue is fiction, so the gate forces LEAD-TIME FICTION regardless of every ratio - and releases the moment intake falls back to or below throughput.
LEAD-TIME FICTIONWhy: On ratios alone the worst read is QUOTING TIGHT - CNC-MILL and ASSY are both cutting it close but still technically covered. But WELD is booking 70 hrs/wk against a demonstrated 48 hrs/wk of throughput even though its own ratio reads COVERED - the backlog there is growing 22 hrs every week. A lead time computed off a growing queue is fiction, so the gate overrides every ratio and forces LEAD-TIME FICTION. 7 work centers - worst ratio-only read QUOTING TIGHT (CNC-MILL, ASSY) - 1 diverging queue (WELD, +22 hrs/wk intake over throughput).
Fix firstWELD - intake is 70 hrs/wk against 48 hrs/wk of demonstrated throughput (+22/wk), booking faster than it can burn
No single work center is the crisis — the clustering is. The roll-up takes the worst, the gate escalates when trouble compounds, and it names the one thing to fix first. Grades the backlog, throughput, and intake numbers you enter - a planning aid, not a scheduling, capacity-planning, or contractual guarantee.
RedHub AI