Production Scheduling for Blow Molding

This guide shows production planners and operations managers how to model blow molding's hybrid-flowshop production in Schantt — from batch blow molders running asymmetric changeovers to continuous downstream stages with staggered shift patterns and multi-route stage skipping for EBM, ISBM, and IBM processes. You will learn how to configure stages, machines, product classes, and calendars, then generate an optimized schedule that minimizes total production time.

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

Industry context

Blow molding produces hollow plastic containers by inflating a heated material against a mold cavity. The industry spans three major process variants — extrusion blow molding (EBM), injection blow molding (IBM), and injection stretch blow molding (ISBM) — each suited to different materials and container geometries. EBM dominates for large, handle-ware, or opaque containers in materials like HDPE and PP, while ISBM and IBM serve clear PET bottles and precision pharmaceutical containers respectively. A typical mid-market plant mixes two or all three processes on the same floor, running products through a shared stage sequence: blow molding, deflashing (EBM only), leak testing, labeling (EBM only), and packaging. Production scheduling in this environment is driven by sequence-dependent changeover times on the blow molders, widely varying throughput rates across product classes, and shift calendars that differ between blowing and packaging.

The blow molder stage is the scheduling bottleneck and the primary source of setup complexity. Changeovers involve swapping mold sets, adjusting temperature profiles, and purging residual material — times range from 15 minutes for a same-material color change to 120 minutes for a full material-type transition. Direction matters: purging dark pigment from a light-colored product takes two to three times longer than the reverse. Plants running 30 to 50 active SKUs across three product classes schedule 12 to 18 changeovers per week, consuming 8 to 12 hours of otherwise productive blowing time. Downstream, the deflashing stations operate at a fixed throughput rate that constrains EBM output under peak load, while packaging runs on a reduced shift pattern that creates a daily buffer of unprocessed stock when blow molders run through the night.

Meridian Blow Molding runs approximately 85 people at a 4,600 m² facility, making 3 product classes across 5 production stages, scheduled by a 2-person planning team.

Process overview

flowchart LR
    BM["Blow Molding<br/>Batch · 6 machines"]
    DF["Deflashing<br/>Flow · 2 machines"]
    LT["Leak Testing<br/>Flow · 2 machines"]
    LB["Labeling<br/>Flow · 1 machine"]
    PK["Packaging<br/>Flow · 2 machines"]
    BM -->|"1 min<br/>conveyor"| DF
    DF -->|"1 min<br/>conveyor"| LT
    LT -->|"1 min<br/>conveyor"| LB
    LB -->|"2 min<br/>conveyor"| PK
    BM -->|"3 min<br/>bridge — PET, PP only"| LT
    LT -->|"2 min<br/>bridge — PET, PP only"| PK

Five-stage production flow through a hybrid batch-and-flow line with multi-route skipping for ISBM and IBM product classes.

PET-clear (ISBM) and PP-natural (IBM) skip Deflashing and Labeling, arriving at Leak Testing via a bridging transfer from Blow Molding, then bridging to Packaging.

Scheduling challenges and how Schantt handles them

In this scenario, the schedule is driven by customer orders for each product class — a fixed set of quantities per week that must be produced and shipped. Your own plant may be driven by a different demand signal, such as make-to-stock targets or a rolling order book; the same configuration principles apply. Schantt builds the schedule forward from a start date and optimizes to minimize total production time — the overall completion time across all jobs — over a practical horizon of one to four weeks. Two scheduling modes are available: Auto mode, where Schantt decides job sequence and machine assignments together from scratch, and Semi-Auto mode, where you provide the job sequence and Schantt optimizes machine assignments within that fixed order.

What Schantt handles well

• Sequential multi-stage production — blow molding's ordered 5-stage route (blow, deflash, leak test, label, pack) with per-class routing and transfer delays between stages, so each downstream step begins only after the upstream step finishes and material arrives.
• Multi-machine stages — 6 blow molders, 2 deflashers, 2 leak testers, and 2 pack stations in parallel; Auto and Semi-Auto modes explore machine assignment across each stage's capable machines to balance load and minimise idle time.
• Mixed batch-and-flow pipelines — blow molding is cyclic batch (clamp, blow, cool, eject per cycle) while deflashing, leak testing, labeling, and packaging run at continuous throughput — both production types coexist in one route and are timed correctly by the simulation.
• Multi-product routing with stage skipping — EBM products pass through all 5 stages; ISBM and IBM products skip deflashing and labeling via bridging transfer times that carry them directly to the next required stage.
• Sequence-dependent changeovers — directional per-machine changeover matrix captures asymmetric purge times for color and material transitions, so the algorithm can favour shorter sequences when building the schedule.
• Shift-aware availability — blow molding runs 126 hours per week (24/5 with Saturday morning) while packaging runs 80 hours per week (2-shift, Monday to Friday), modeled via stage-specific calendars that clamp operations to their actual working windows.

How Schantt handles each challenge

1. Asymmetric color and material changeovers consuming 8 to 12 hours per week.

• Changeover duration depends on both the material pairing and the direction of the transition. A dark-to-light color purge takes up to 50 minutes, while light-to-dark takes only 20 minutes. Material-type transitions to and from PET require a full temperature re-profile lasting 120 minutes. With 12 to 18 changeovers across 6 blow molders each week, sequencing the wrong direction compounds the time penalty.
• Schantt models each changeover as a directional per-machine entry — the setup time from one product class to another on a specific blow molder, where the from-to value can differ from the to-from value. When the system evaluates candidate schedules, each changeover is added to the operation's start time and therefore to total production time. In Auto mode, Schantt explores job sequences across all blow molders to find a combination that minimizes the cumulative changeover penalty; in Semi-Auto mode, it optimizes machine assignments within your fixed order.

2. Multi-cavity rate variation of 15 times across the product mix.

• PET-clear runs at approximately 18,000 units per hour (12 cavities, fast ISBM cycle), while HDPE-opaque 1-litre bottles run at approximately 1,029 units per hour (4 cavities, 14-second cycle). PP-natural sits in between at approximately 2,160 units per hour. Assuming uniform throughput misallocates blow-molder capacity, overbooking time for PET runs and under-reserving for HDPE runs.
• Each product class on the blow molder stage has its own batch size (number of cavities) and cycle duration. Schantt uses these parameters to calculate the true duration of each job: the total quantity divided by the batch size, multiplied by the cycle time. Downstream flow stages convert throughput directly from the rate. The schedule accounts for each class's actual production rate, so capacity is allocated proportionally and the plan reflects real run times.

3. Parallel machine assignment across 6 blow molders.

• The 6 blow molders share the same product-class capabilities but each job must be assigned to one specific molder at a time. With 30-plus SKUs and frequent order changes, manually assigning each job to a machine and sequencing those assignments to minimize changeover penalties is the most time-consuming planning decision.
• Schantt treats each blow molder as an independent resource within the same stage. In Auto mode, the scheduling algorithm explores both job-to-machine assignments and the running order on each machine, choosing the combination that minimizes overall completion time. In Semi-Auto mode, the job sequence is fixed and Schantt optimises which machine each operation runs on. The resulting assignment appears on each operation in the schedule and Gantt view.

4. Deflashing constraining EBM throughput under peak load.

• Two deflashers serve six blow molders, a 1-to-3 ratio. Under normal conditions this is adequate, but when multiple HDPE-opaque jobs run simultaneously on high-cavity molds, deflashing throughput of 1,800 units per hour per station becomes the bottleneck. Blow molders idle on wait-material back-pressure while deflashers are at capacity, limiting HDPE-opaque output to roughly 85 percent of blow-molder capacity during peak periods.
• Deflashing is modeled as a flow stage with two parallel machines, each with a throughput rate of 1,800 units per hour for HDPE-opaque product. The schedule simulation feeds each deflashing station from the blow molder that precedes it and applies supply-chain timing — when accumulated material exceeds what deflashing can clear, the downstream station queues work while upstream operations show wait-material segments. The planner sees the bottleneck as a timing delay on the Gantt rather than discovering it on the floor.

5. Staggered shift patterns creating daily buffer accumulation.

• Blow molding runs 126 hours per week (24 hours Monday to Friday plus Saturday morning), while packaging runs only 80 hours per week (2 shifts, Monday to Friday). The 46-hour gap means blow molders produce through the night while packaging stands idle. By Friday evening the packaging bay is congested, and by Monday morning blow molders cannot start new product because previous runs remain unprocessed.
• Schantt assigns a separate calendar to the packaging stage — 06:00 to 22:00, Monday to Friday — while blow molding, deflashing, leak testing, and labeling use the production calendar (24/5 with Saturday morning). Schedule simulation clamps packaging operations to its calendar windows only, so work scheduled outside those hours is held until the next shift opens. The Gantt shows packaging paused during non-working hours as shaded overlays, while blow molders continue overnight for as long as their calendar permits.

What to model in Schantt

The following entities cover the production setup for this blow-molding scenario:

Step-by-step setup

1. Create the stages and set transfer times. Create 5 stages in order: Blow molding (batch type), Deflashing (flow), Leak testing (flow), Labeling (flow), and Packaging (flow). On each stage's detail page, enter the transfer time to the next consecutive stage — 1 minute from Blow to Deflash, 1 minute from Deflash to Leak, 1 minute from Leak to Label, and 2 minutes from Label to Pack. For the skip routes, add bridging transfers: 3 minutes from Blow directly to Leak, and 2 minutes from Leak directly to Pack.

2. Add the machines to each stage. Create 13 machines and assign each to its stage:

• Blow molding (6): BM-01, BM-02, BM-03, BM-04, BM-05, BM-06
• Deflashing (2): DF-01, DF-02
• Leak testing (2): LT-01, LT-02
• Labeling (1): LB-01
• Packaging (2): PK-01, PK-02

3. Create the product classes and define routing. Create 3 product classes: HDPE-opaque, PET-clear, and PP-natural. On each class's detail page, set per-class routing — HDPE-opaque passes through all 5 stages in order; PET-clear and PP-natural route through Blow molding, then Leak testing (skipping Deflashing and Labeling), then Packaging. Leave partial transfers disabled for all legs.

4. Add one product per class. Create one representative product within each class: Motor Oil 1L Black (HDPE-opaque), Water 500ml (PET-clear), and Pharmaceutical Drops 250ml (PP-natural). Each product inherits its class's routing, machine capability, and processing parameters automatically.

5. Set machine capacity parameters and changeovers on each blow molder. On each of BM-01 through BM-06, enter the batch parameters for each product class the machine supports:

• HDPE-opaque: batch size = 4 units per cycle, cycle time = 0.23 minutes
• PET-clear: batch size = 12 units per cycle, cycle time = 0.04 minutes
• PP-natural: batch size = 6 units per cycle, cycle time = 0.17 minutes

On the flow stages, set throughput per product class per machine:

• Deflashing (DF-01, DF-02): 1,800 units per hour, HDPE-opaque only
• Leak testing (LT-01, LT-02): 2,500 units per hour, all classes
• Labeling (LB-01): 1,500 units per hour, HDPE-opaque only
• Packaging (PK-01, PK-02): 1,200 units per hour, all classes

Then configure the directional changeover matrix on each blow molder. For each of BM-01 through BM-06, enter 6 per-pair entries — one for every from-to combination among the 3 product classes:

• HDPE-opaque to PET-clear: 120 minutes (and reverse: 120 minutes)
• HDPE-opaque to PP-natural: 82 minutes (and reverse: 82 minutes)
• PET-clear to PP-natural: 120 minutes (and reverse: 120 minutes)

6. Configure calendars, exceptions, and downtimes. Create two calendars. Set the production calendar (126 hours per week, 24 hours Monday to Friday plus 00:00 to 06:00 Saturday) as the team default — it applies to blow molders, deflashers, leak testers, and the labeler by default. Create a second packaging calendar (80 hours per week, 06:00 to 22:00 Monday to Friday) and assign it to PK-01 and PK-02. Add 2 calendar exceptions for team-wide non-working days: January 1 (New Year's Day) and May 1 (International Workers' Day). Finally, add 3 machine downtimes: a factory-wide year-end shutdown from December 24 to December 31, quarterly preventive maintenance on BM-03 (January 8, 14:00 to 18:00), and trim-blade replacement on DF-01 (March 12, 08:00 to 12:00).

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

Common mistakes

1. A single blanket changeover instead of a directional per-pair matrix. Entering one changeover time for all transitions on a blow molder ignores the 3-to-1 asymmetry between dark-to-light and light-to-dark purges. The scheduling algorithm treats each transition as equally time-consuming and will not favour the shorter direction. Fix: Enter every from-to pairing with its own duration — 120 minutes for PET transitions, 82 minutes for HDPE-PP swaps, and within-class color pairs if your plant runs multiple colours per material.

2. Merging deflashing and labeling into the blow molder stage. Rolling deflashing and labeling into a single stage prevents the model from reflecting that ISBM and IBM products skip those steps. The schedule would show PET and PP jobs occupying Deflashing and Labeling capacity — or worse, requiring capacity that does not exist on the floor. Fix: Keep Deflashing, Leak testing, Labeling, and Packaging as separate stages, and set per-class routing to skip the stages each product class does not use.

3. Assuming uniform throughput across all product classes. Using the same throughput or cycle time for every class on the blow molder stage misallocates capacity. HDPE-opaque runs at roughly one-fifteenth the rate of PET-clear, and a schedule built from a single rate will overbook some machines and under-load others. Fix: Enter distinct batch parameters per product class on each blow molder — cavity count as batch size and per-cycle time as cycle duration — as well as distinct throughput rates on each flow-stage machine.

4. Forgetting the packaging calendar gap. The blow molding production calendar and the packaging calendar differ by 46 hours per week. If packaging machines use the same 126-hour calendar as the blow molders, the schedule will show packaging running through the night and weekend — work that no one will be there to do. Fix: Assign the packaging-specific calendar (2-shift, Monday to Friday) to PK-01 and PK-02, and leave the default production calendar for all other machines.

5. Configuring all three process variants as a single product class. EBM, ISBM, and IBM products follow different routes through the plant. Grouping them into one class forces every product through all 5 stages — including deflashing and labeling steps that ISBM and IBM bottles never visit — or, conversely, prevents EBM products from reaching those stages at all. Fix: Create a separate product class for each process variant (HDPE-opaque for EBM, PET-clear for ISBM, PP-natural for IBM) and set each class's routing to match its true floor path.

What a good schedule looks like

With Schantt configured for the 5-stage line, 13 machines, 3 product classes, directional changeovers, and separate production and packaging calendars, the difference over a representative week is evident.

Before (manual Gantt planning):

• Changeovers are sequenced without a system to compare directional pairs, so dark-to-light transitions appear as often as the reverse — adding 30 to 40 minutes per swap compared to the optimal direction.
• Blow molder assignments are made by habit rather than by load balance, leaving some molders idle while others queue jobs behind a long changeover.
• The packaging shift gap is tracked in the planner's head; by Thursday, blown stock has accumulated beyond the packing stations' clearing capacity, and Friday's schedule requires weekend overtime to catch up.
• Deflashing bottlenecks appear only when the morning operator reports a conveyor backup, forcing an emergency reprioritisation of the EBM runs.

After (Schantt Auto mode):

• The directional changeover matrix guides the algorithm toward sequences that favour shorter transitions, reducing cumulative changeover time by dozens of minutes per week without manual pairing analysis.
• Blow molder workload is distributed across all 6 machines according to each class's true run rate — HDPE-opaque runs on the appropriate number of molders while PET-clear occupies fewer machines for shorter calendar time.
• The production and packaging calendars are both active in the simulation: packaging starts each shift working through the stock built up overnight, and the schedule respects its 22:00 end time automatically. Buffer accumulation falls within the packing stations' clearing capacity.
• Deflashing is modelled as a constrained flow stage; when multiple EBM jobs approach peak throughput, the simulation delays upstream blow molder starts just enough to prevent idle deflashers from being overwhelmed, and wait-material segments on the Gantt make the constraint visible before it causes floor disruption.