Production Scheduling for Pressure-Sensitive Adhesives

Production planners and plant managers at pressure-sensitive adhesive coating facilities scheduling multiple coating lines, slitters, and sheeters across differing chemistries can model the full production pipeline in Schantt — from coating through converting — and let the optimizer find a sequence that reduces total changeover time.

This guide follows a fictional composite company built from industry research on pressure-sensitive adhesives; all names, parameters, and figures are illustrative.

Industry context

Pressure-sensitive adhesive (PSA) manufacturing turns liquid adhesive into coated, dried, and converted finished goods for packaging, labelling, and industrial assembly. The process begins on a coating line where liquid adhesive — hot-melt styrenic block copolymer, solvent-borne acrylic, or water-borne emulsion — is applied to a moving substrate (BOPP film, paper facestock, or tissue carrier) at a controlled coat weight. The coated web passes through a heated drying or curing tunnel where solvent evaporates or the hot-melt cools, then rewinds into a jumbo roll up to 1.5 metres wide. Those jumbo rolls intermediate between coating and converting: they cool, residual solvent dissipates, and they await transfer to downstream machines in a staged buffer.

Converting splits into two distinct paths. Tape products (permanent carton-sealing and double-coated) move to slitting, where wide jumbo rolls are cut into narrow finished rolls on razor or shear slitters. Label-stock products skip slitting entirely and go to sheeting, where the web is cut into rectangular sheets, counted, banded, and palletised. A lamination stage sits between coating and slitting for double-coated products only, where a second substrate is combined with the coated web under heat and pressure.

Coating throughput varies substantially by chemistry: hot-melt products coat fastest at 6,000 m²/hr, solvent acrylic at 5,400 m²/hr, and high-tack formulations at 4,800 m²/hr. Changeovers between chemistry families require adhesive-pan drain, solvent purge of the delivery system, die or roll swap, and line re-prime — durations from one to several hours depending on which chemistry is leaving and which is arriving. Directional asymmetry is the norm: moving from a dark-colour product to a light-colour one requires a full cleanout because visible residue would compromise the next batch, while the reverse direction may skip the full purge.

Mid-market PSA facilities typically operate two to four coating lines, each capable of a subset of coating methods (slot-die, comma-coat) and chemistries. Lines are not interchangeable — method exclusivity and chemistry compatibility mean that each line has a defined capability envelope. Converting machines (slitters and sheeters) operate on different shift patterns from coating, often running two shifts while coating runs continuously. This calendar mismatch creates a coordination problem: coating produces jumbo rolls through the night and weekend, but converting capacity only starts absorbing them the next morning.

Crestview Coated Products runs approximately 95 people at a 5,000 m² single-site facility, making three product classes across four production stages, scheduled by a two-person planning team.

Process overview

flowchart LR
    N["Coating (Flow) — CL1, CL2, CL3"]
    N2["Lamination (Flow) — L1"]
    N3["Slitting (Batch) — S1, S2"]
    N4["Sheeting (Batch) — SH1, SH2"]
    N -->|"15 min"| N2
    N -->|"120 min"| N3
    N -->|"120 min"| N4
    N2 -->|"60 min"| N3

Four-stage production flow — three coating lines feed a laminator, two slitters, or two sheeters depending on product class. Transfer times between stages reflect jumbo-roll handling, curing and cooling dwell, and crane transport.

DblCoatTape routes Coating → Lamination → Slitting. PermTape skips lamination and routes Coating → Slitting. RemLabel skips lamination and slitting, routing Coating → Sheeting. Bridge transfer times from coating to slitting (120 min) and coating to sheeting (120 min) cover the skipped curing and cooling dwell and handling.

Scheduling challenges and how Schantt handles them

The scenario assumes a hybrid make-to-order and make-to-stock demand profile — approximately 60% cycle-driven base load (standard products run on a weekly fixed cycle for stock replenishment) with 40% order-driven top-up (custom widths, specialty adhesives, short-run orders) — over a two-week rolling horizon. (Readers whose operation is purely make-to-order or purely make-to-stock will still find the model applicable; the horizon is configurable in Schantt.) The scheduling algorithm minimises total production time across all stages — the time from the first job starting on coating through the last finished roll or pallet exiting converting — scheduling forward from a start date. Customer order due dates serve as context for sequencing decisions but are not hard-enforced constraints; the planner reconciles delivery commitments at schedule review. Auto mode optimises both job sequence and machine assignment across parallel lines, exploring combinations that a human planner could not evaluate manually. Semi-Auto mode lets the planner fix the sequence while Schantt optimises machine assignment — useful when customer priorities, raw-material availability, or liner stock dictate the order and only the line assignment remains open.

What Schantt handles well

• Sequence-dependent changeovers. Directional changeover time pairs on coating lines capture chemistry-based cleanout durations and directional asymmetry — for example, a solvent-to-hot-melt switch might be 120 minutes while the reverse is 90 minutes. The optimizer naturally clusters compatible classes to minimise total changeover time across the plant.
• Multi-machine stages with capability-restricted machines. Three coating lines, two slitters, two sheeters — each with defined capability sets encoded through per-class throughput entries. A coating line that cannot run solvent chemistry simply has no throughput entry for that class. Auto and Semi-Auto modes assign jobs to eligible machines while respecting these implicit constraints.
• Multi-product routing with stage skipping. PermTape (coat then slit), RemLabel (coat then sheet), DblCoatTape (coat then laminate then slit) — three divergent routes in one model. Bridging transfer times handle skip-span handoffs.
• Mixed batch-and-flow pipelines. Coating and lamination as continuous flow stages; slitting and sheeting as batch stages with cycle time and batch size. Partial transfer at converting allows downstream to begin on the first usable portion of a jumbo roll.
• Shift-aware availability. Diverse calendars — 24/7 coating line, 24/5 lines, 5×8 lamination, 5×16 slitting, staggered morning and afternoon sheeter shifts. Machine-level calendar overrides model each line's actual working hours.
• Per-class throughput differentiation. Different chemistries run at different line speeds (PermTape 6,000 m²/hr, RemLabel 5,400 m²/hr, DblCoatTape 4,800 m²/hr). Schantt schedules duration proportionally rather than assuming a uniform line speed.

How Schantt handles each challenge

1. Chemistry-driven changeover time on coating lines.
- Coating lines require extensive cleanout when switching between chemistry families — adhesive-pan drain, solvent purge, die swap, and re-prime. The longest crossing (acrylic emulsion to hot-melt SBC) takes several hours; lighter crossings within hot-melt families take about one hour. Directional asymmetry is significant. With ten to fifteen changeovers per week across three lines, Crestview loses eight to fourteen hours of coating capacity to changeover each week.
- Schantt captures each coating line's directional changeover times as paired entries — one for each from→to direction, with unequal values where asymmetry exists. The optimizer evaluates the total changeover penalty of each candidate sequence and naturally prefers sequences that cluster compatible chemistries. A sequence that groups all hot-melt PermTape runs then all hot-melt DblCoatTape runs incurs only the one-hour intra-family changeover rather than the multi-hour cross-family purge, and the optimizer finds that clustering automatically.

2. Coating-line assignment with chemistry constraints.
- Not every coating line can run every chemistry. CL1 is hot-melt only (PermTape, DblCoatTape). CL2 handles solvent and water-borne only (RemLabel, DblCoatTape). CL3 can run both hot-melt and solvent (PermTape, DblCoatTape). RemLabel is locked to CL2 — if CL2 is down, there is no backup for label-stock. The planning team manually resolves this three-way assignment weekly.
- Schantt encodes each coating line's capability through which product classes have throughput entries on that machine. A machine without a rate entry for a given class is not a candidate for that work. When the planner loads the order book, Auto mode evaluates all eligible machines for each job and selects the combination that minimises total production time across the plant. If CL2 is unavailable due to maintenance, the schedule immediately reflects that RemLabel cannot run — no hidden assumption that a backup line exists.

3. Slitter and sheeter changeover accumulation at volume.
- Crestview's two slitters run five sixteen-hour days, processing both PermTape and DblCoatTape in mixed sequences. Width-pattern changeovers (five to twenty minutes each) accumulate: with typical batch quantities of ten to thirty jumbo rolls per product, Crestview averages eight to twelve changeovers per slitter per day. Across both slitters and both sheeters, converting capacity loss to setup is twenty to thirty hours per week.
- Schantt models slitter and sheeter changeovers as per-machine, per-direction duration pairs, just as on coating lines. The optimizer sequences converting jobs to minimise the number of wide-to-narrow or narrow-to-wide transitions and clusters same-width runs together. The same model captures sheeter format-change times (three to five minutes per size switch), so the schedule balances changeover load across all four converting machines.

4. Divergent routing and stage coordination.
- Three product classes, three routes, two converting destinations, one shared coating stage. PermTape and DblCoatTape compete for slitter time. RemLabel requires only sheeting — a separate machine set on different shifts. The laminator (day shift only) gates DblCoatTape throughput. The planner must coordinate the order in which coating produces each class, how long jumbo rolls wait (minimum sixty-minute dwell plus handling), and when converting slots open.
- Schantt models each product class's routing independently — PermTape passes through coating then slitting, RemLabel through coating then sheeting, DblCoatTape through coating then lamination then slitting. Transfer times set on the stage detail page define the minimum handoff delay between stages (coating to slitting: 120 minutes; coating to sheeting: 120 minutes; coating to lamination: 15 minutes; lamination to slitting: 60 minutes). Partial transfer allowances on slitting and sheeting let converting begin on the first five hundred square metres of a jumbo roll rather than waiting for the full roll to complete, reducing idle time between stages. The schedule resolves all four routes simultaneously, so a coating run that serves all three classes sequences the coating jobs in an order that keeps each downstream machine fed.

What to model in Schantt

The following first-class entities form the core of the PSA model in Schantt.

Step-by-step setup

1. Create the stages and set transfer times. Create four stages in process order. Set each stage's production type: Coating (flow, position 10), Lamination (flow, position 30), Slitting (batch, position 50), Sheeting (batch, position 60). On the Coating stage detail page, add transfer times to each downstream stage that any product class reaches:

Coating to Lamination: 15 minutes — direct hall transfer between adjacent machine areas
Coating to Slitting: 120 minutes — bridge transfer covering cooling dwell, residual solvent evaporation, roll handling, and crane transport
Coating to Sheeting: 120 minutes — same bridge logic for the label-stock route

On the Lamination stage detail page, add a 60-minute transfer time to Slitting for roll handling after lamination.

2. Add the machines to each stage. Add eight machines to their respective stages. Assign each machine its calendar at creation:

Coating: CL1 (slot-die, 24/7 calendar), CL2 (comma-coat, 24/5 calendar), CL3 (slot-die, 24/5 calendar)
Lamination: L1 (5×8 day shift calendar)
Slitting: S1 (5×16 double shift calendar), S2 (5×16 double shift calendar)
Sheeting: SH1 (5×8 morning shift calendar), SH2 (5×8 afternoon shift calendar)

3. Create the product classes and define per-class routing. Create three product classes with square metres as the scheduling unit: PermTape, RemLabel, DblCoatTape. On each class's detail page, add the stages that class passes through:

PermTape: Coating (no partial transfer) → Slitting (partial transfer allowed, quantity 500 m²)
RemLabel: Coating (no partial transfer) → Sheeting (partial transfer allowed, quantity 500 m²)
DblCoatTape: Coating (no partial transfer) → Lamination (no partial transfer) → Slitting (partial transfer allowed, quantity 500 m²)

Partial transfer on slitting and sheeting means the first 500 m² of a coated jumbo roll can move to converting before the full roll is finished, reducing idle time between stages.

4. Add one product per class. Create CT-48 (PermTape, 48 mm × 1,000 m rolls), AL-210 (RemLabel, A4 sheets), and DT-12 (DblCoatTape, 12 mm × 50 m rolls). Each product inherits its class's routing automatically — you configure routing once per class, not per product.

5. Set machine capacity and changeovers. On each coating line's detail page, enter throughput values — the line speed in square metres per hour — for only the product classes that line can run:

CL1: PermTape 6,000 m²/hr, DblCoatTape 4,800 m²/hr
CL2: RemLabel 5,400 m²/hr, DblCoatTape 4,800 m²/hr
CL3: PermTape 6,000 m²/hr, DblCoatTape 4,800 m²/hr

On L1, enter DblCoatTape throughput at 4,800 m²/hr. On S1 and S2, enter batch cycle time (3 minutes per roll) and batch size (1,000 m²) for both PermTape and DblCoatTape. On SH1, enter batch cycle time (30 minutes per batch) and batch size (1,000 sheets) for RemLabel; on SH2, enter 40-minute cycle at the same batch size (SH2 runs larger sheet formats at a slower rate).

Then add changeover time pairs on every machine that runs more than one product class:

CL1: PermTape to DblCoatTape 60 min, DblCoatTape to PermTape 60 min
CL2: RemLabel to DblCoatTape 90 min, DblCoatTape to RemLabel 120 min
CL3: PermTape to DblCoatTape 60 min, DblCoatTape to PermTape 60 min
S1 and S2: PermTape to DblCoatTape 10 min, DblCoatTape to PermTape 10 min

The unequal CL2 values (90 min vs. 120 min) capture directional asymmetry: moving from acrylic emulsion to high-tack acrylic takes less cleanout than the reverse due to residual polymer build-up. Setting both directions explicitly is how Schantt models this asymmetry — not through a sequencing rule, but through duration pairs that make the optimizer naturally prefer the direction with less changeover time.

6. Configure calendars, exceptions, and downtimes. Set the default calendar to Monday through Friday, 24-hour production. Override each machine with its specific calendar. Add three team-wide calendar exceptions: New Year's Day (Jan 1), International Workers' Day (May 1), and Christmas Day (Dec 25) — all non-working. Add machine downtimes: CL1 monthly deep clean (second Sunday each month, 06:00–10:00, maintenance), S1 weekly preventive maintenance (Friday 14:00–15:00, maintenance), and year-end plant shutdown (Dec 26 06:00 – Dec 31 18:00, all machines).

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

Common mistakes

1. Using a single blanket changeover instead of per-pair directional times. A single changeover value applied to all coating-line transitions hides the real scheduling constraint. The optimizer cannot distinguish a one-hour hot-melt-to-hot-melt switch from a two-hour solvent-to-hot-melt purge, so it has no incentive to cluster compatible chemistries — it treats all transitions as equal. Fix: Enter each directional pair with its actual duration. On CL2, for example, RemLabel to DblCoatTape is 90 minutes, but the reverse is 120 minutes; both directions need their own entry. The optimizer will then prefer sequences that minimise total changeover time across the line.

2. Defining one product class for all routed products. A single class covering both tape and label-stock forces all products through the same routing — either slitting or sheeting, but not both. The schedule cannot model the divergent paths that define PSA converting, and the planner must manually override every job's stage list. Fix: Create separate product classes for each distinct routing pattern. Three classes (PermTape, RemLabel, DblCoatTape) let each product follow its real path through the plant — coating then slitting, coating then sheeting, or coating then lamination then slitting — without per-job overrides.

3. Setting all coating lines to the same throughput. If every coating line uses a single throughput value, the schedule treats a thirty-minute PermTape run and a thirty-minute DblCoatTape run as identical in duration — but DblCoatTape coats at 4,800 m²/hr versus PermTape at 6,000 m²/hr, a twenty percent difference. Over a two-week horizon the error compounds significantly. Fix: Enter per-class throughput for each coating line matching its actual line speed for that chemistry.

4. Forgetting to bridge skipped stages with transfer times. When RemLabel routes from Coating directly to Sheeting (skipping Lamination and Slitting), the schedule needs a transfer time for that skip-span handoff. Without it, the default zero-minute transfer tells the optimizer that a jumbo roll can reach the sheeter instantly, which is false — it still needs cooling dwell and handling. Fix: Set a 120-minute transfer time from Coating to Sheeting and from Coating to Slitting to cover the real-world delay regardless of which intermediate stages are skipped.

What a good schedule looks like

A well-configured PSA schedule coordinates coating and converting across three divergent routes, sequences changeovers to cluster compatible chemistries, and respects each machine's calendar constraints. The difference between Crestview's manual weekly planning process and a Schantt optimised schedule shows in the operational metrics.

Before (manual spreadsheet):

• Planner assigns PermTape permanently to CL1, DblCoatTape to CL3, and RemLabel to CL2 — no cross-loading even when one line is idle and another is queued, because manually tracking capability across three lines is impractical
• Coating changeovers sequenced by planner intuition, often mixing solvent and hot-melt runs on the same day — incurring the full multi-hour purge two or three times per shift instead of clustering compatible chemistry families
• Slitter changeovers scheduled reactively as jobs arrive from coating, with no attempt to cluster same-width PermTape runs — eight to twelve changeovers per slitter per day, each five to twenty minutes
• Stage coordination is manual and fragile: the planner estimates when coating will finish each class, leaves buffer for dwell and handling, and manually schedules converting slots — any disruption forces a full recalculation that takes two to three hours

After (Schantt Auto mode):

• Coating jobs automatically assigned to eligible lines — CL1 and CL3 share PermTape and DblCoatTape based on current load, freeing capacity on both and reserving CL2 exclusively for RemLabel without planner intervention
• Compatible hot-melt runs clustered: all PermTape followed by all DblCoatTape on each line, incurring only the one-hour intra-family changeover between them rather than a multi-hour cross-family purge — changeover time consolidated into fewer, shorter transitions
• Slitter changeovers grouped by width pattern — same-width PermTape runs batched together across S1 and S2, reducing daily setup time on each machine from hours to minutes
• Transfer times and partial-transfer allowances synchronise coating and converting automatically; the sheeter starts processing RemLabel sheets within minutes of the first partial transfer arriving, rather than waiting for the full jumbo roll to cool and be moved — idle time between stages drops from hours to minutes

The result is a schedule that runs the same order book in measurably less total production time — not because the machines run faster, but because changeover time is consolidated, machine assignments flex with demand, and every stage stays fed by the stage before it without planner intervention for every job transition.