Production Scheduling for Tire Manufacturing

Production planners and operations managers at mid-market tire plants can use Schantt to schedule a hybrid flowshop with batch mixing, flow calendering and extrusion, batch tyre building, batch curing, and flow finishing — modelling multi-machine stages, sequence-dependent changeovers, and shift-aware availability — then load the built-in example dataset to explore the configuration in minutes.

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

Industry context

Tire manufacturing is a linear hybrid-flowshop process that transforms raw rubber, carbon black, oils, and reinforcing materials into finished tyres through seven sequential stages. Each stage runs multiple parallel machines, and product classes — passenger radial, light truck, and high-performance — share the same equipment while differing in compound, dimensions, and processing parameters.

The process begins at Mixing, where a Banbury mixer blends the rubber compound in batch cycles. Passenger radial (PR) compound mixes at 220 kg per batch in 9.5-minute cycles, light truck (LT) at 260 kg in 11.5-minute cycles, and high-performance (HP) compound at 150 kg in 10-minute cycles. High-performance compound is segregated on a dedicated mixing line to avoid cross-contamination. The mixed compound moves to Calendering, where fabric and steel cord are coated with rubber at throughputs of 350–450 units per hour depending on class, and to Extrusion, where tread and sidewall profiles are formed at 250–300 units per hour. A dedicated extruder handles high-performance compound. Meanwhile, Bead and Apex Building produces bead wire assemblies at 0.4–1.0 kg per batch in 0.9–1.6-minute cycles. These supply streams converge at Tyre Building, where components are assembled into green tyres on building drums at 9–15 tyres per batch in 1.33–2.33-minute cycles.

From tyre building, green tyres transfer to Curing — the bottleneck stage — where 25 presses mould and vulcanise each tyre under heat and pressure. Cure cycles range from 10–14 minutes (PR), 16–22 minutes (LT), and 9–13 minutes (HP). Each mould change between product classes takes 30–50 minutes. After curing, tyres move through Finishing and Inspection, where three inspection lines run at 50–65 units per hour per line, performing visual checks, X-ray inspection, and uniformity testing.

Vellum Tyre Company runs 360 people at a 48,000 m² facility, producing three product classes across seven production stages, scheduled by a five-person planning team. The plant operates 2 mixers, 1 calender, 2 extruders, 2 bead winders, 12 building drums, 25 curing presses, and 3 inspection lines. Shift patterns vary by department: mixing runs 24/5, building runs 20/5, curing runs mixed 24/5 and 24/7, and finishing runs 16/5.

Process overview

flowchart LR
  M["Mixing<br/>(batch)"] --> C["Calendering<br/>(flow)"]
  C --> E["Extrusion<br/>(flow)"]
  E --> BB["Bead and Apex Building<br/>(batch)"]
  BB --> TB["Tyre Building<br/>(batch)"]
  TB --> CU["Curing<br/>(batch)"]
  CU --> FI["Finishing and Inspection<br/>(flow)"]

Each product class flows through all seven stages in a single linear sequence — mixing, calendering, extrusion, bead and apex building, tyre building, curing, and finishing and inspection. Curing is the bottleneck stage.

Schantt schedules each product class — passenger radial, light truck, and high-performance — as a single linear sequence through all seven stages, from mixing to finishing and inspection. A physical plant prepares calendered liner/ply, extruded tread/sidewall, and beads in parallel before they meet at tyre building, but Schantt models calendering, extrusion, and bead-and-apex-building as sequential stage positions, not converging branches; it does not schedule a bill-of-materials merge.

Scheduling challenges and how Schantt handles them

In this scenario the schedule is driven by a fixed demand plan — a list of product-and-quantity jobs for each tire class over a one-week horizon. (Plants whose primary driver differs, such as make-to-order, can still follow the same configuration approach while adjusting the job list.) The optimizer minimizes total production time, scheduling all jobs forward from a chosen start date. Schantt offers two optimisation modes: Auto mode, where the algorithm decides job sequence, machine assignments, and timing from scratch; and Semi-Auto mode, where the planner fixes the production order and the algorithm optimises machine assignments and detailed timing within that sequence.

What Schantt handles well

• Sequential multi-stage production — Tire manufacturing is a linear hybrid flowshop. Schantt models the ordered stage sequence once, assigns each product class its own routing, and chains downstream stages to start only after upstream stages finish plus any configured transfer time.
• Multi-machine stages — Every tire stage runs multiple parallel machines. Schantt explores machine assignments across each stage, restricting to capable machines only, and shows the assignment on the Gantt.
• Mixed batch-and-flow pipelines — Tire routes mix batch physics (mixing, building, curing) with flow physics (calendering, extrusion, finishing). Schantt types each stage batch or flow, derives timing from the right parameters, and simulates supply-driven starvation visible as wait-material segments on the Gantt.
• Sequence-dependent changeovers — Mold changes, purges, die changes, roll-profile changes, and size changeovers all depend on the from-to product-class pair. Schantt's directional changeover matrix captures per-machine, per-pair durations; Auto mode favours lower-changeover sequences.
• Shift-aware availability — Mixing runs 24/5, building runs 20/5, curing runs 24/5 or 24/7, and finishing runs 16/5. Schantt assigns separate calendars per stage, clamps starts into working windows, and advances by working time only.
• Calendar exceptions and downtimes — Holidays, maintenance shutdowns, and year-end shutdowns affect different calendar groups. Team-wide calendar exceptions override every calendar on a given date; machine downtimes subtract capacity before scheduling.

How Schantt handles each challenge

1. Curing-press bottleneck and mold-change overhead.

• Cure cycles last 7–15 times longer than tyre-building cycles, making curing the clear bottleneck. A set of 25 presses handles output from 12 building drums, and each mould change between product classes takes 30–50 minutes of lost production time. With multiple class changes per week, cumulative changeover overhead on the curing stage can reach 8–15 hours. A planner scheduling by spreadsheet or tribal knowledge has no systematic way to minimise mould-change frequency or visualise its impact on throughput.
• Schantt models each curing press with per-class batch parameters and a directional changeover matrix that captures every from-to mould-change duration. In Auto mode, the algorithm sequences jobs to cluster same-class runs together where that reduces total mould-change time, then assigns each curing job to the press whose calendar and availability best suit the schedule. The resulting changeover segments appear as labelled bars on the Gantt ahead of each operation's processing bar, so the planner can see — and validate — exactly how much time each transition consumes. (Because Schantt does not track individual mould inventory, the planner must independently confirm that the correct mould is available on the assigned press for each scheduled job.)

2. Shift-calendar mismatch and the green-tire bank.

• Mixing and calendering run 24/5 (three rotating crews), building runs 20/5 (two 10-hour shifts), curing is split between 24/5 and 24/7 presses, and finishing runs 16/5 (two 8-hour shifts). These mismatched calendars create an intentional buffer — a bank of green tyres waiting for cure — but also mean that a building shift ending at 02:00 Friday cannot feed curing presses that run through the weekend if the building crew is idle. Without a schedule that respects each department's working windows, the planner must manually track which shifts feed which presses.
• Schantt assigns a separate working calendar to each stage (and, for curing, a different calendar to Press 2 versus Press 1). All start times are clamped into the relevant working window, and durations advance by working time only — skipping non-working gaps automatically. On the Gantt, non-working stretches appear as shaded overlays, so the planner sees exactly when each stage is available and can confirm that the green-tire gap between building completion and curing start falls within each class's cure-window limit (48–72 hours for PR and LT, 8–24 hours for HP).

3. Compound segregation requiring dedicated lines.

• High-performance compound must not cross-contaminate with passenger or light-truck compounds. The plant dedicates Mixer B and Extruder B to HP compound exclusively, with no purge between those machines and other classes. A schedule that ignores this restriction risks assigning HP jobs to the wrong mixer or extruder, introducing contamination.
• Schantt models compound segregation through per-machine rate-entry restriction: Mixer B has processing parameters only for the HP product class, and Extruder B has throughput entries only for HP. Because the algorithm can assign a job only to a machine that has a valid duration entry for that product class and stage, HP jobs are locked to their dedicated equipment. Mixer A and Extruder A handle PR and LT classes only, with purge changeovers of 8 minutes (mixer) and 50 minutes (extruder) between class transitions.

4. Multiple changeover types across different stages.

• A tire plant does not have one kind of changeover — it has mixer purges (8 minutes), calender roll-profile changes (60–75 minutes), extruder die changes (50 minutes), bead-winder size changeovers (15–20 minutes), builder drum size changeovers (30–45 minutes), and press mould changes (30–50 minutes), each with asymmetric durations between class pairs. A planning tool that applies a single blanket changeover time per machine or ignores directionality significantly misrepresents the actual time penalty of switching classes.
• Schantt captures each machine's changeovers as a directional per-pair matrix — the duration from one product class to another on a specific machine, where from-to can differ from to-from. The calender, both bead winders, both builders, and both curing presses all carry a full six-pair matrix. The algorithm folds the correct changeover duration into each operation's start time, so schedules that cluster similar classes to avoid time-consuming transitions score better in total production time.

5. Multi-shift material handoff with partial transfer.

• High-performance green tyres have a tight cure window of 8–24 hours, meaning they must reach a curing press promptly after building. A standard full-batch handoff — wait for the entire batch to finish at tyre building before transferring — risks pushing the first-built tyres past their viable cure window. The plant needs the curing stage to begin consuming tyres as soon as the first units are ready, without waiting for the full building batch to complete.
• Schantt supports partial transfer at the tyre-building-to-curing handoff for the HP class, with a partial-transfer quantity of 1 tyre. This lets the downstream press begin curing the first tyre while the building drum is still assembling the remainder of the batch. The configuration is set on the HP product class's routing at the tyre-building stage. All other product classes and handoffs use full-batch transfer. On the Gantt, the planner sees overlapping processing bars at tyre building and curing for HP jobs, confirming that the cure window is being respected — though the final verification that each green tyre's gap stays within its class limit remains a manual Gantt check.

What to model in Schantt

The following table lists the first-class entities a planner creates to model this tire manufacturing scenario. Sub-configuration — per-class routings, changeovers, transfer times, calendar exceptions, and machine downtimes — is set on each entity's detail page.

The dataset models 14 machines as a representative subset of the plant's 47-floor total. The 2 building drums represent 12 floor drums, and the 2 curing presses represent 25 floor presses — each shares identical processing parameters and changeover behaviour, so the reduction preserves parallel-machine and changeover dynamics while keeping the configuration manageable.

Step-by-step setup

1. Create the seven stages in order. Add Mixing (batch), Calendering (flow), Extrusion (flow), Bead and Apex Building (batch), Tyre Building (batch), Curing (batch), and Finishing and Inspection (flow) at positions 1 through 7. On each stage's detail page, set the transfer times to the next stage:

• Mixing to Calendering: 5 minutes
• Calendering to Extrusion: 8 minutes
• Extrusion to Bead and Apex Building: 12 minutes
• Bead and Apex Building to Tyre Building: 15 minutes
• Tyre Building to Curing: 10 minutes
• Curing to Finishing and Inspection: 40 minutes

2. Add the machines to each stage. Add 2 mixers to Mixing, 1 calender to Calendering, 2 extruders to Extrusion, 2 bead winders to Bead and Apex Building, 2 building drums to Tyre Building, 2 curing presses to Curing, and 3 inspection lines to Finishing and Inspection. Curing Press 2 uses the 24/7 calendar override; all other machines use stage-level calendar assignment.

3. Create the three product classes and define each class's routing. Create Passenger Radial, Light Truck, and High-Performance. Each class routes through all seven stages in sequence. For the High-Performance class, enable partial transfer at the Tyre Building stage with a quantity of 1, and verify the allow-partial-transfer toggle is set.

4. Add one representative product per class. Add 205/55R16 (Passenger Radial), 265/70R17 (Light Truck), and 255/35R19 (High-Performance), each with a distinct display colour for Gantt visibility.

5. Set each machine's processing parameters and changeovers. On each machine's detail page, enter the per-class parameters:

• Mixer A: PR 220 kg batch, 9.5 min cycle; LT 260 kg batch, 11.5 min cycle. Purge changeover PR↔LT: 8 min each way.
• Mixer B: HP 150 kg batch, 10 min cycle only — no changeovers needed.
• Calender 1: PR 450, LT 350, HP 400 units/hr. Roll-profile changeovers: all six directional pairs, 60–75 min.
• Extruder A: PR 300, LT 250 units/hr. Die change PR↔LT: 50 min each way.
• Extruder B: HP 280 units/hr only — no changeovers.
• Bead Winder 1 & 2: All three classes. Size changeovers: all six directional pairs, 15–20 min.
• Builder 1 & 2: PR 10 tyres, 1.33 min cycle; LT 15 tyres, 2.33 min cycle; HP 9 tyres, 1.50 min cycle. Drum changeovers: all six directional pairs, 30–45 min.
• Press 1 & 2: PR 10 tyres, 12 min cycle; LT 15 tyres, 18 min cycle; HP 9 tyres, 11 min cycle. Mould changeovers: all six directional pairs, 30–50 min, asymmetric.
• Inspection Lines 1–3: PR 65, LT 50, HP 60 units/hr — no changeovers.

6. Configure calendars, exceptions, and downtimes. Set four calendars — 24/5 Continuous (mixing, calendering, extrusion, Press 1), 20/5 Two-Shift Building (bead building, tyre building), 24/7 Continuous (Press 2), and 16/5 Day Shifts (finishing and inspection). Add six team-wide calendar exceptions for non-working days. Add three machine downtimes: Press 1 annual preventive maintenance, a year-end curing shutdown, and Mixer A rotor rebuild.

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

Common mistakes

1. Using a single blanket changeover time per machine instead of directional per-pair values. Curing-press mould changes are asymmetric — 30 minutes from PR to LT but 40 minutes from LT to PR. A single value misrepresents the real time penalty and leads the algorithm to underestimate or overestimate transition time depending on sequence direction. Fix: Enter both directions for every product-class pair that shares a machine, using the durations from the dataset.

2. Assigning the same calendar to all stages. When mixing, building, curing, and finishing each run different shift patterns, a single calendar causes the algorithm to schedule operations outside actual working windows. Fix: Create the four separate calendars described in the setup and assign each stage (or individual machine, for Press 2) to its correct calendar.

3. Creating one product class that covers divergent compound-handling requirements. If Passenger Radial and High-Performance compounds are in the same product class, both would share the same routing and machine assignments, and HP jobs could be placed on the shared mixer. Fix: Separate classes by compound segregation requirements and set per-machine processing parameters only for the classes each machine is allowed to run.

4. Omitting partial transfer for time-sensitive handoffs. Without partial transfer enabled at the tyre-building-to-curing handoff for high-performance tyres, the curing press waits for the entire building batch to finish before starting — potentially pushing the first-built units past their 8–24-hour cure window. Fix: On the High-Performance class routing at the Tyre Building stage, enable partial transfer with a quantity of 1.

5. Entering machine counts that do not reflect the actual number of parallel resources. If the model shows 25 curing presses but only 2 are entered, the curing stage appears vastly more constrained than it is — or vice versa. Fix: Match the machine count for each stage to the dataset counts specified in the entity table, which preserve the plant's per-stage parallel-machine dynamics.

What a good schedule looks like

When the model is configured correctly, the difference between a manual spreadsheet approach and an optimised Schantt schedule is visible in the Gantt and the total production time.

Before (manual scheduling): The planning team sequences jobs by tribal knowledge, with no systematic mould-change minimisation. Curing-press availability is tracked on paper or in ad-hoc spreadsheets, and green-tire cure-window gaps are checked manually and inconsistently. Symptoms include:

• Curing-press utilisation varies widely across the week, with some presses idle while others queue
• Mould-change sequences are arbitrary, consuming 8–15 hours of cumulative changeover time per week
• Green-tire dwell times for high-performance jobs occasionally exceed the 24-hour limit, discovered only when the tyres reach curing
• Shift mismatches between building and curing cause unplanned idle periods that compound delays into the finishing shift

After (Schantt Auto mode): With the full configuration in place and a week's demand as the job list, the algorithm produces a complete schedule in minutes. The Gantt shows:

• Curing-press operations clustered by product class, with mould-change overhead visibly reduced through sequence optimisation
• Each press assigned according to its calendar — Press 1 on 24/5, Press 2 on 24/7 — with no job scheduled outside its working windows
• Each press's cure-cycle and changeover parameters are visible, so the planner can visually confirm that green-tire gaps between building completion and curing start fall within each class's cure-window limit (48–72 hours for PR and LT, 8–24 hours for HP)
• Partial-transfer overlap for HP jobs, with curing beginning on the first tyre before the building batch finishes
• Non-working periods, maintenance downtimes, and calendar exceptions rendered as shaded overlay bands, explaining every pause in the production flow

The planner can review the result, adjust the sequence or parameters, re-run in Auto mode, or switch to Semi-Auto mode to lock a preferred order while letting the algorithm assign machines and timing.