Mould-bound plants live and die on the press log.
Injection-moulding plants are mould-bound. Throughput is governed by which mould runs which press for which shift, how often the press swaps colour or material, and how cycle-time drift on each mould-press pair erodes OEE between OEE reports. Polymr keeps the 8-week mould-slot plan synced to live demand, ingests cycle-time drift directly from the press logs, and surfaces regrind feedstock supply against virgin spend before the variance hits the P&L.
Built for the plastics shape.
Polymr parses mould-design CAD (STEP, IGES, SLDPRT) into press-cycle-time estimates per cavity, regrind feedstock requirements per shift, and the changeover envelope each mould needs to swap colour or material. Tool-trial data and the press log feed back into the same model so the cycle-time estimate corrects to the actual press-mould pair within the first 200 shots.
Color changeover scheduling reads the ERP back-to-back conflicts before the planner sees them. A black-after-natural run blocks until the purge cycle fits the available window. Mould cycle-cap tracking compares the press-log shot count against the finance cycle-cap reserve, and the discrepancy lands on a single row before the toolmaker quote is due.
Six failure modes that govern a mould-bound plant.
Each is what a plant manager will name when asked why mould utilization is not 85 percent. None of these are about press tonnage or material spec as individual choices; each is about the seam between the press log, the ERP schedule, the regrind loop, and the toolmaker where today no workflow runs.
- Mould cycle-cap mid-quarterMould M-202 hits annual cycle cap mid-quarter; replacement tooling not in next-quarter forecast
Mould M-202 was specified for 800k cycles per year. Demand pulled forward, the mould ran 540k cycles in the first 22 weeks, and the cycle cap arrives in W34 not W52. Replacement tooling has a 14 to 16 week lead from the toolmaker. Nobody flagged it because the cycle counter lives in the press log and the mould-life model lives in a finance spreadsheet.
- Regrind feedstock dropsIn-house regrind feed for PMR-PL-44 drops; virgin resin spend spikes 14 percent
The regrind loop for PMR-PL-44 feeds 18 percent of new shots. A downstream OEM raised reject thresholds on the part, reducing the recoverable scrap pool by half. Regrind feed drops from 18 percent to 9 percent. Virgin resin demand spikes 14 percent with a 6-week lag before procurement re-baselines, and the unit cost on the next 800k shots moves three points.
- Two SOs collide on the same mouldSO-3318 and SO-3344 both need mould M-211 in W22; no automated re-peg
Sales accepts SO-3344 for 80k units in W22 without seeing that M-211 is already committed to SO-3318 for 96k units in the same window. The collision surfaces at the Monday standup. Replanning ripples through three downstream SOs because the mould-slot view lives on a printed sheet on the planner's desk.
- Cycle-time drift undetectedMould M-208 cycle drifts from 38s to 44s over 6 weeks; throughput shortfall surfaces W7
Press 2 running M-208 was at 38s per cycle in W19. By W22 it crept to 41s. By W25 it was 44s. The OEE report runs weekly and rolled-up; the cycle-time signal lives in the MES at the cell-level. The press operator did not see the trend because the operator dashboard rolls daily. The throughput shortfall surfaces on the W25 fulfilment report.
- Changeover scheduled back-to-backP-12 clear to P-44 black needs 4hr cleaning; ERP schedules them back-to-back
Product P-12 runs in clear resin; product P-44 runs in black. The colour changeover requires a 4-hour purge-and-clean cycle (loss of 36 shots per minute equivalent). The ERP scheduler places P-44 immediately after P-12 because it does not carry product-pair colour-changeover times. Press 3 loses an entire shift to the purge that was schedulable around.
- Tooling repair delays nine SOsM-209 tooling repair takes M-209 out for 9 days; 9 SOs delay; expedite spend $42K
Mould M-209 needs a tooling repair (worn cavity insert). The toolmaker quotes 9 days. Nine SOs that peg to M-209 in W23-W24 need replanning. Half re-peg to M-211 with cycle-time delta; half need expedited shipping on alternate suppliers. The total expedite spend lands at $42K for the quarter because the replan happened five days after the repair quote landed.
The machine schedule the press operator and the planner read off the same row.
Press 2 today. Seven slots, two product families, one changeover slot, one maintenance window. The changeover slot CO-99 is the 4-hour P-12 clear to P-44 black purge the ERP previously placed back-to-back. The maintenance window is the cavity-insert clean cycle the press operator schedules around the next-shift run. The schedule is the single artifact the press operator and the planner both read; no printed sheet on the planner's desk.
| Slot | WO | SKU | Mould | Cycle | Qty | Status |
|---|---|---|---|---|---|---|
| 07:00 | WO-3421 | P-12 clear cap | M-204 | 38s | 1,800 | run |
| 09:30 | WO-3422 | P-12 clear cap | M-204 | 38s | 1,200 | run |
| 11:30 | CO-99 | Changeover P-12 → P-44 | M-204→M-208 | 4 h | . | changeover |
| 15:30 | WO-3426 | P-44 black lid | M-208 | 41s | 2,200 | run |
| 19:00 | WO-3427 | P-44 black lid | M-208 | 41s | 1,400 | queued |
| 22:00 | MAINT | Insert clean cycle | M-208 | 45 min | . | maint |
| 23:00 | WO-3431 | P-44 black lid | M-208 | 41s | 1,800 | queued |
The cycle-drift on M-208 surfaces at +0.4s, not at +6s.
Press 2 running M-208 was at 38.0s per cycle in W17. By W19 it had crept to 38.4s. By W25 it was 44.2s. The OEE report rolls weekly and the operator dashboard rolls daily; neither catches the drift inside two weeks of the baseline crossing.
Polymr ingests the cycle signal from the press log (MES API or direct OPC-UA depending on press vintage), maintains the per-mould-per-press baseline against the prior 12 weeks, and flags drift the first time the trailing-shift average crosses the +0.4s tolerance band. The 6-second drift that today surfaces on the W25 fulfilment report becomes a Tuesday afternoon planning conversation in W19.
The mould master row carries the cycle counter, the refurb gate, and the toolmaker repair quote.
M-209 is 180 cycles away from the 8,000-cycle refurb gate and currently out of service on a 9-day tooling repair. The cycle counter blocks the press at the refurb gate; the toolmaker repair quote attaches to the row rather than the planner's inbox; the fallback mould (M-211, +1.4s per cycle) is the row the scheduler reads when the M-209 SO bucket re-pegs.
- Mould ID
- M-209
- Type
- 2-cavity, hot runner
- Press fit
- Press 3 (220t), Press 4 (250t)
- Cycle baseline
- 46.0s
- Cavity geometry
- PMR-PL-44 hex insert
- Cycles to date
- 7,820 / 8,000 refurb
- Cycles cap
- 720,000 / year
- Status
- Tooling repair, ETA 9 d
- Toolmaker
- V-742 (Akron Tool & Die)
- Fallback
- M-211 (4-cavity, +1.4s)
- Last service
- 2026-04-18 . insert swap
- Next refurb gate
- 180 cycles
What this looked like at a two-site injection-moulding plant.
- Situation
- Each site ran 22 moulds across three presses on a three-shift schedule. Mould-changeover, cycle-time variance, and resin colour batching governed how a week of demand became a week of running schedule.
- What was breaking
- MRP ran against yesterday's inventory. Planners burned 9-12 hours per week reconciling on-hand scans, open POs, and in-process scrap before they trusted output. Mould utilization peaked below 71% because changeover windows were planned manually around the wrong demand picture.
- Planning + purchasing
- Margin and bottleneck analysis
- Engineering revisions
Three workflow families do the heavy lifting in a mould-bound plant.
- Planning + purchasing
Net requirements refresh against live on-hand and live press-cell scrap. Recommendations sized around mould + press constraints; colour-changeover cost rolled into the slot allocation so the schedule does not place P-12 clear immediately before P-44 black.
- Margin + bottleneck + revisions
Press-level OEE, cycle-time drift attribution per mould-press pair, colour-changeover cost surfaced into the weekly plan. Mould revisions (cavity insert change, gating geometry update) propagate to affected open POs and the changeover plan.
- Delay recovery
Virgin resin slips, regrind feedstock drops, and toolmaker delays consolidate to a single decision packet. The expedite spend that today fragments across the planner, the buyer, and the toolmaker coordinator consolidates with the trade-offs ranked.
Three siblings that share the mould-bound shape.
- High-volume
Cycle-time-driven cousin. The takt-driven planning patterns that hold a high-volume automotive line at fill rate apply directly to a plastics plant when the press cycle is the bottleneck.
- Process
Lot and yield neighbour. The regrind feedstock loop on a plastics plant shares the lot-genealogy and yield-variance shape of a process plant; the recipe substitution and release-gate patterns transfer.
- Tires & rubber
Mould-cured neighbour. Industrial tire and rubber goods share the mould-utilization-and-cure-time bottleneck shape; the same schedule-and-changeover workflow applies.