Production Scheduling for Injection Molding

This guide is for production planners, operations managers, and plant managers in custom injection molding who want to model their press floor in Schantt — capturing directional changeover matrices, multi-cavity batch rates, per-class routing, and shift-aware calendars so the schedule reflects the real physics of injection molding production.

This guide follows a fictional composite company built from industry research on injection molding; all names, parameters, and figures are illustrative.

Industry context

Injection molding transforms thermoplastic and thermoset resins into finished parts by injecting molten material into a clamped mold under high pressure. The press floor runs a mix of electric and hydraulic machines in different tonnage classes — small presses for high-volume commodity parts, mid-range presses for engineered components, and large presses for oversized or deep-draw parts. Each press is a self-contained production cell with dedicated auxiliary equipment: hopper loader, temperature controller, mold heater or chiller, and often a robot or sprue picker. Material-specific drying is done at the press — nylon 66, for instance, requires four hours of drying before processing — so each press's readiness depends on its upstream material preparation being complete before the mold closes. Changeovers between product runs are sequence-dependent: switching from a dark, glass-filled material to a natural resin requires a full barrel purge that can take an hour, while the reverse transition is much faster.

A typical custom molding facility runs hundreds of active SKUs that route through a shared press floor before diverging into secondary operations. Some parts need only degating and packaging — the runner system is snapped off and the parts are bagged. Others require manual trimming of flash or gate vestige before packing. Hot-runner tools produce no runner waste and can skip degating entirely. Changeover time at an injection press is not a single number: switching from a natural PP to a dark ABS on the same press takes roughly 30 minutes, but the reverse transition — dark ABS back to natural PP — requires a full barrel purge that can consume 55 to 75 minutes. The same asymmetry applies when moving between cold-runner and hot-runner material classes: stabilising a hot-runner manifold after switching from a cold-runner material adds roughly 20 minutes beyond the material-purge time alone. With roughly 35 changeovers every week across the press floor, these directional differences accumulate into hours of hidden capacity that no flat estimate can capture.

The planning challenge is to sequence jobs across parallel presses of differing tonnages so that changeover time is minimized, downstream stages are fed at the right rate, each product class visits only the stages its process requires, and time-sensitive customer orders are not sacrificed to changeover efficiency.

NexForm Plastics Corp. runs approximately 85 people at a single 4,200 m² facility, producing around 400 active SKUs organized into roughly 30 product classes across three representative families: a high-volume polypropylene cold-runner part for consumer packaging, a glass-filled Nylon 66 hot-runner part for automotive, and a medium-volume ABS cold-runner part for electronics enclosures. The plant operates 24/5 from Monday 06:00 through Saturday 06:00 on three rotating 8-hour shifts. A three-person planning team schedules approximately 35 changeovers per week across the press floor, currently using a flat 30-minute changeover estimate for every transition.

Process overview

flowchart LR
    M["Molding<br/>(batch — 8 presses)"]
    D["Degating<br/>(flow — 4 stations)"]
    T["Trimming<br/>(flow — 2 stations)"]
    P["Packaging<br/>(flow — 3 stations)"]

    M -->|"Cold-runner<br/>(PP, ABS)"| D
    M -->|"Hot-runner<br/>(Nylon66)"| P
    D -->|"Trim-required<br/>(ABS)"| T
    D -->|"No trim needed<br/>(PP)"| P
    T -->|"Trimmed parts<br/>(ABS)"| P

Parts flow from the press floor through degating and optional trimming before converging at packaging. Hot-runner parts bypass degating entirely.

The dataset uses per-class routing so each product class visits only the stages its process requires. The hot-runner class (glass-filled Nylon 66) skips degating and trimming — no runner waste and no flash. The ABS class uses all four stages. The PP cold-runner class skips trimming.

Scheduling challenges and how Schantt handles them

The schedule at a custom molding facility is driven by incoming customer orders and blanket purchase orders. (If your operation is make-to-stock, the same configuration applies — your trigger is replenishment rather than order intake.) Schantt schedules forward from a chosen start date and minimizes total production time, the time from the first job's start to the last job's completion across all stages. A practical planning horizon for this scenario is two to four weeks, balancing order visibility with schedule stability.

Schantt offers two scheduling modes for this scenario. Auto mode optimises both the job sequence and the machine assignments, exploring re-orderings that cluster similar changeovers and balance press loads. Semi-Auto mode preserves your fixed production sequence while optimising which machines each job runs on — ideal when the order is operationally locked but machine selection is not.

What Schantt handles well

• Directional changeover matrices. Configure per-press changeover times between every product-class pair, with different durations for each direction. The algorithm favours job sequences that cluster similar material-and-color classes to reduce purge time.

• Multi-cavity batch rate calculation. Each press × product-class combination carries its own cavity count and cycle duration. Schantt derives the correct production rate from these batch parameters rather than relying on a fixed machine-throughput figure.

• Per-class routing with stage skipping. Each product class follows exactly the stages it needs. A hot-runner class skips degating entirely; a trim-requiring cold-runner class includes trimming. Shared presses and transfer times bridge across skipped stages.

• Shift-aware and exception calendars. Calendars define working shifts (24/5) with date-level exceptions for holidays and plant shutdowns. Press-level downtime windows block maintenance and breakdown periods, and jobs route around them.

• Auto and Semi-Auto scheduling modes. Auto mode sequences jobs and assigns machines to minimise total production time. Semi-Auto mode preserves a planner-determined job sequence while optimising machine assignments, with earliest-start constraints available for time-sensitive jobs.

How Schantt handles each challenge

1. Sequence-dependent changeover times.

• Changeover time at an injection press depends on both the material leaving and the material coming in, and the difference can be dramatic — at NexForm, the actual weighted average across 35 weekly changeovers is 47 minutes, not the 30-minute flat estimate the team uses. A dark-glass to natural-polymer transition can consume nearly an hour of purge time while the reverse takes a quarter of that. Using a single changeover number for all transitions systematically misstates capacity.
• Schantt models changeover times as a directional per-press matrix — a different duration for every product-class pair in each direction. On a single press shared by three material classes, nine directional pairs define every transition. The algorithm's Auto mode naturally favours sequences that group similar classes (natural→natural, dark→dark) to reduce purge time, and the resulting time penalty is visible on the Gantt as a labelled segment before each job's processing bar.

2. Mixed cavity counts and cycle durations across parallel presses.

• A 16-cavity PP cap mold running an 18-second cycle on a 90-ton press produces at a very different rate than a 2-cavity ABS backplate mold running a 28-second cycle on a 400-ton press. Treating every press-and-product combination as having the same throughput distorts load balance and misleads delivery commitments.
• Schantt models each press and product-class combination with its own batch parameters: cavity count and cycle time. The nominal production rate is derived from the number of cycles needed to produce the ordered quantity at the stated cavity count, so each job's duration reflects the real tool-and-press combination. The simulation charges exactly those durations and feeds them into machine availability, so downstream stages see accurate arrival times.

3. Divergent per-class routings with stage skipping.

• Three product classes that share the same press floor follow different paths after molding. The hot-runner automotive connector skips degating and trimming entirely — the hot runner eliminates runner waste and there is no flash to remove. The PP bottle cap needs degating but no trimming. The ABS electronics backplate needs both. Scheduling these side by side on a whiteboard makes downstream WIP timing nearly invisible.
• Schantt assigns each product class its own routing as an ordered set of stages. The hot-runner class visits only Molding and Packaging; the PP class visits Molding, Degating, and Packaging; the ABS class visits all four stages. Transfer times bridge across skipped stages — a direct Molding→Packaging transfer of 15 minutes handles the hot-runner skip — so no stage is silently missing a handoff. On the Gantt, each product appears only on the stages its class requires, and the material flow is visible from press to final packout.

4. Unplanned downtimes and calendar exceptions that derail the plan.

• A missed 24-hour annual maintenance window on the 1,000-ton press (PL-03) can invalidate an entire week's schedule if discovered mid-week. An 8-hour heater-band breakdown on a 90-ton press forces a last-minute reassignment of the PP cap order. Three plant holidays and a year-end shutdown further reduce available hours, and without a formal calendar the planner must remember each one.
• Schantt models plant-wide calendar exceptions (holidays, shutdowns) and machine-specific downtime windows (maintenance, breakdowns) directly. Both are subtracted from working capacity before any timing is calculated, so jobs never land inside a blocked window. When a downtime is entered on the machine detail page, the schedule routes around it automatically, and the blocked intervals render as shaded Gantt overlays with the reason and category visible on hover.

5. Time-sensitive customer orders competing with changeover-efficient sequencing.

• Automotive connector housings have firm delivery commitments — 8% of automotive orders shipped late in the past quarter, averaging 2.3 days late, most traced to a sequencing strategy that prioritised changeover minimisation over customer timing. The planner needs to keep high-priority jobs early in the week without reverting to fully manual scheduling.
• Semi-Auto mode lets the planner fix the production sequence — urgent automotive jobs ordered first — while Schantt optimises machine assignments within that sequence. Per-job earliest-start constraints push the automotive connector housing jobs to begin as early as possible on their first stage. The algorithm then assigns each job to the best available press without reordering, trading changeover time against machine load within the planner's priority frame.

What to model in Schantt

The following five first-class entities form the configuration surface for this injection molding scenario. Sub-configuration such as changeover matrices, transfer times, batch parameters, and throughputs are set on each entity's detail page.

Step-by-step setup

Configure Schantt in this order so each entity's dependencies are in place before you need them.

1. Create the stages. Create four stages in position order: Molding (batch, position 10), Degating (flow, position 20), Trimming (flow, position 30), Packaging (flow, position 60). On each Stage detail page, set the transfer times between consecutive stages and across skip routes:

• Molding → Degating: 5 minutes
• Molding → Packaging: 15 minutes (skip-bridge for hot-runner class)
• Degating → Trimming: 2 minutes
• Degating → Packaging: 5 minutes
• Trimming → Packaging: 2 minutes

2. Add machines to each stage. On the Molding stage, add 8 presses covering the facility's tonnage range. On Degating, add 4 stations (3 manual benches and one pneumatic press). On Trimming, add 2 stations (one pneumatic press and one manual bench). On Packaging, add 3 stations (one automated count-and-bag line and two manual packing stations).

3. Create product classes and define routings. Create three product classes, each with a distinct per-class routing set on its detail page:

• PP-Natural-ColdRunner routes through Molding → Degating → Packaging
• Nylon66-GF30-Black-HotRunner routes through Molding → Packaging (skips Degating and Trimming)
• ABS-Dark-ColdRunner routes through Molding → Degating → Trimming → Packaging

Verify that each class's routing includes only the stages its products actually visit. No partial-transfer legs are needed for this dataset — each job completes at one stage before the next begins.

4. Add one product per class. Add a representative product for each class: PP Bottle Cap 28mm for the PP cold-runner class, Automotive Connector Housing for the hot-runner class, and Electronics Enclosure Backplate for the ABS class. Each product inherits its class's routing and machine capability.

5. Configure machine parameters and changeovers. On each press, enter batch parameters per product class — cavity count (batch size) and cycle duration — so Schantt derives the correct production rate per run:

Batch parameters — Molding press × product class:
- PP-Natural-ColdRunner on PS-02, PS-03, PM-01, PM-03: 16 cavities, 0.30 min cycle
- ABS-Dark-ColdRunner on PM-01, PM-03, PM-04, PL-01: 2 cavities, 0.47 min cycle
- Nylon66-GF30-Black-HotRunner on PM-03, PM-04, PM-06: 4 cavities, 0.58 min cycle

Then set throughput values on each downstream station (Degating, Trimming, Packaging). On Degating, manual benches run cold-runner parts at 1,800 units per hour across both PP and ABS classes, while the pneumatic press runs at 7,200 units per hour — four times faster, making it the preferred station when it is available. On Trimming, the pneumatic press handles ABS parts at 450 units per hour and the manual bench at 240 units per hour. On Packaging, the automated count-and-bag line runs at 90,000 units per hour across all three classes, while manual stations run at 900 units per hour each — the automated line alone handles the entire PP bottle cap volume while manual stations absorb lower-rate automotive and electronics orders.

Finally, on each shared press, enter the directional changeover matrix for every product-class pair the press runs. For a press running all three classes, this produces 9 directional pairs — natural→natural at 25 minutes, natural→ABS at 30 minutes, natural→nylon at 75 minutes, and so on in each direction. The asymmetry (dark→natural at 55–75 minutes versus natural→dark at 10–30 minutes) is captured by the directional entry.

6. Configure calendars and downtimes. Create the Standard 24/5 calendar (Monday 06:00 to Saturday 06:00) and add four date-level exceptions: New Year's Day, International Workers' Day, Christmas Day, and a year-end shutdown on December 24. Optionally, add three machine downtime entries — the 1,000-ton press annual preventive maintenance window (24 hours in July), a mold maintenance window on the 220-ton press (8 hours), and a breakdown repair on the 90-ton press (8 hours). These are subtracted from working capacity automatically, and jobs route around them.

For step-by-step instructions on configuring each of these in Schantt, see the Schantt documentation.

Common mistakes

1. Entering one flat changeover time for all transitions on a shared press. A single changeover duration makes every transition look the same, but a dark-glass to natural-polymer purge takes nearly three times as long as the reverse. The schedule systematically underestimates press idle time and overstates available capacity. Fix: Enter per-pair directional changeover times so each transition carries its real duration.

2. Using a fixed throughput rate on the Molding stage instead of batch parameters. Entering a single units-per-hour figure on a press ignores the fact that cavity count and cycle duration vary by product class. A 16-cavity PP mold running an 18-second cycle produces at a completely different rate than a 2-cavity ABS mold. Fix: Set Molding as a batch stage and enter cavity count and cycle duration for each product class the press runs.

3. Consolidating hot-runner and cold-runner products under one product class. A single product class forces all its products through the same routing. When some products use a hot runner (no runner waste, no flash) and others use a cold runner (needs degating), one group will always be wrong. Fix: Create separate product classes for hot-runner and cold-runner workflows, each with its own per-class routing.

4. Omitting skip-bridge transfer times for routes that bypass an interior stage. Defining transfer times only between consecutive stages leaves a hot-runner part with no transfer time from Molding to Packaging — the handoff defaults to zero, which is never realistic. Fix: Add a direct Molding→Packaging transfer time of 15 minutes to bridge the skipped Degating and Trimming stages.

5. Starting a schedule without confirming mold availability. All the press and changeover configuration in the world does not help if the mold is still on another press, past its maintenance interval, or not yet clean. A schedule built on assumed mold availability is at risk of infeasibility. Fix: Before committing a schedule, verify each mold's status — location, shot count, and cleanliness — against the facility's tracking system.

What a good schedule looks like

The difference between a flat-estimate whiteboard schedule and a Schantt-optimised plan appears in five measurable areas.

Before (flat changeover estimates, whiteboard sequencing): Approximately 10 hours per week of hidden press idle time caused by the gap between the 30-minute flat changeover assumption and the actual 47-minute weighted average — roughly 1.5 presses of capacity lost to nothing but mismeasured setup time. Hot-runner and cold-runner jobs interleaved arbitrarily, incurring unplanned 20-minute-plus purge premiums every time the runner type switched. Multi-stage WIP — what is sitting at degating, what is waiting for trimming — invisible until someone walks the floor, producing 4 to 8 hours per week of latent WIP accumulation. The July maintenance window on the 1,000-ton press and the PS-02 heater-band breakdown creating a persistent gap between planned and actual capacity that the planner discovers only when the week is already in motion. Eight percent of automotive connector housing orders shipping late by an average of 2.3 days, five of the nine late orders traced directly to changeover-only sequencing that ignored customer timing needs.

After (Schantt Auto + Semi-Auto): Changeover time variance captured directionally on each shared press, recovering the approximately 10 hours per week of press capacity that the flat estimate was hiding — every shared press now reflects the real purge profile of its material pairs. Runner-type and colour groups naturally clustered by the algorithm, reducing unnecessary purge changeovers — cross-runner swaps minimised, same-class sequences preferred, and the purge time premium paid only where physically unavoidable. Multi-stage WIP visible on the Gantt as each product class follows its correct stage path with defined transfer times, so the degating queue and trimming workload are no longer a blind spot — downstream operators see what is arriving before it reaches their station. Calendar exceptions and machine downtime entries enforced automatically — the July PM window on the 1,000-ton press and the August breakdown on the 90-ton press are blocked out and jobs route around them, ending the gap between planned and actual capacity. Automotive connector housing jobs sequenced earlier in the week via Semi-Auto mode with earliest-start constraints, bringing the 8% late-order rate toward zero and giving the planning team a repeatable method for protecting time-sensitive work without resorting to fully manual scheduling.