Production Scheduling for Gummy Supplements

This guide is written for production planners and plant managers in gummy supplement manufacturing. It shows how to configure Schantt for a hybrid flowshop with batch cooking, parallel depositing lines, per-class drying dwells, sequence-dependent changeovers, and three packaging lines — then run optimised schedules that respect every constraint.

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

Industry context

Gummy supplement production is a batch-driven hybrid flowshop. The process begins with cooking: gelatin, sweeteners, active ingredients, and colourants are combined in steam-heated kettles to form a liquid slurry. Each kettle runs a 1,000 kg batch with a cycle duration of approximately 45 minutes. The slurry is then deposited into starch moulds across six parallel depositing machines, where per-class batch sizes range from 160 kg to 200 kg with cycle durations of 18–22 minutes. After depositing, the moulds enter a drying hold that varies by product type — 8 hours for clear gummies, 18 hours for opaque, and 12 hours for coated — before being demoulded, oiled for gloss, and sent to packaging. Coated gummies undergo an additional sugar-polish stage in two coating drums before packaging, while clear and opaque gummies bypass coating entirely.

The changeover landscape adds significant sequencing complexity. Kettle changeovers between different colour or formulation classes range from 15 to 25 minutes depending on the direction. Depositing machine cleanouts are even more demanding: switching from a light clear formulation to a dark opaque formulation requires thorough cleaning that can take 60–90 minutes. Demould stations carry 8–20 minute changeovers, and packaging lines require 45 minutes per transition. An unoptimised schedule can lose 3–6 hours of productive time per day to changeovers alone.

NutriGummy Co. runs approximately 90 people at a 6-stage, 25-machine facility, making three product classes — Clear (gelatin-based), Opaque (high-solids), and Coated (sugar-polish) — across 2 shifts (Monday to Friday, 06:00–22:00), scheduled by a 2-person planning team.

Process overview

flowchart LR
  Cooking["Cooking"] --> Depositing["Depositing"]
  Depositing --> Demoulding["Demoulding"]
  Demoulding --> Oiling["Oiling"]
  Demoulding --> Coating["Coating"]
  Oiling --> Packaging["Packaging"]
  Coating --> Packaging

Six-stage gummy supplement production flow: batch stages (Cooking, Depositing, Demoulding, Oiling, Coating) feed into a flow stage (Packaging). The drying dwell between Depositing and Demoulding is modelled as a per-class transfer time, not a dedicated stage.

Skip-routing note: The coated product class skips Oiling and routes through Coating instead. Clear and opaque classes route through Oiling and skip Coating.

Scheduling challenges and how Schantt handles them

Gummy supplement scheduling is driven by a demand forecast of product quantities by class over a weekly or biweekly horizon — in this scenario, a weekly demand plan across the three classes. (If your demand arrives as daily orders or a monthly push, the same schedule structure applies; you may prefer Semi-Auto mode to lock in the run order.) The scheduling algorithm minimises total production time — the overall completion time across all jobs — by finding the best job sequence and machine assignments. Schantt schedules forward from a start date and, for this guide, assumes a 2-week practical horizon. The two optimisation modes are Auto (the system decides both sequence and machine assignments) and Semi-Auto (you provide the production order, and the system optimises machine assignments within it).

What Schantt handles well

• Sequential multi-stage production with transfer times — The full 6-stage gummy flow (Cooking, Depositing, Demoulding, Oiling, Coating, Packaging) with per-class drying dwell as transfer times between Depositing and Demoulding.
• Multi-machine stages with parallel assignment — Cooking kettles, depositing machines, demould stations, oiling tumblers, coating drums, and packaging lines each as parallel-machine pools that the scheduler assigns automatically.
• Mixed batch-and-flow pipelines — Batch stages (Cooking, Depositing, Demoulding, Oiling, Coating) and a flow stage (Packaging) in the same route, each with its own duration physics.
• Multi-product routing with stage skipping — Three product classes with divergent routings: the coated class skips Oiling and routes through Coating instead, using per-class skip-routing.
• Sequence-dependent changeovers — Colour- and flavour-based cleanouts on depositing machines and kettles modelled as directional per-machine changeovers; the scheduler can reorder jobs to cluster similar colours.
• Shift-aware availability with calendar exceptions and downtimes — The two-shift pattern (Monday to Friday, 06:00–22:00) modelled as a calendar; holidays and maintenance windows as calendar exceptions and machine downtimes.

How Schantt handles each challenge

1. Long drying dwell after depositing.

• Gummies must dry in a holding room after depositing before they can be demoulded. Clear gummies require approximately 8 hours, opaque gummies approximately 18 hours, and coated gummies approximately 12 hours. This multi-hour hold is orders of magnitude longer than any other operation on the line — a cooking cycle is 45 minutes and a depositing cycle is 18–22 minutes — and it forces the schedule to wait while the room fills. The drying room itself has finite rack positions that are not tracked, so the planner must monitor capacity separately.
• Schantt models the drying dwell as a per-class transfer time from Depositing to Demoulding, set to the required drying hours in minutes. Because transfer time is wall-clock elapsed time, the hold runs continuously across shifts and weekends — a product that finishes depositing at 16:00 begins demoulding after the full 8, 12, or 18 hours of elapsed wall time, not after only working minutes. The schedule chains each product's demoulding start to begin only after its drying dwell has elapsed, without requiring a standby machine for the hold itself. If production volumes approach the drying-room rack limit, the planner confirms capacity manually since room occupancy is not part of the model.

2. Sequence-dependent colour changeovers on depositing machines.

• Switching from a light clear batch to a dark opaque batch on a depositing machine requires thorough cleaning that can take 60–90 minutes, depending on the direction. An unoptimised weekly schedule can lose 3–6 hours of productive time per day just to these cleanouts. Kettle changeovers (15–25 minutes) and demould station changeovers (8–20 minutes) add further sequencing constraints that compound across the full production run.
• Schantt models each changeover as a directional per-machine time from every product class to every other. When the scheduler evaluates candidate job sequences, it adds the changeover time between consecutive jobs to each operation's start time and therefore to total production time. In Auto mode, the algorithm can reorder jobs to cluster similar colour classes on each depositing machine, reducing the total changeover penalty. In Semi-Auto mode, the planner fixes the run order and the system optimises machine assignments within it, still respecting the sequence-dependent durations — depositing machines that receive a cleaner switch get the darker product first, for example.

3. Divergent per-class routings with stage skipping.

• Clear and opaque gummies pass through Oiling and skip Coating, while coated gummies skip Oiling and route through Coating instead. These divergent paths share the same upstream stages (Cooking, Depositing, Demoulding) and the same final stage (Packaging), so the schedule must interleave products correctly across the split after Demoulding and the merge into Packaging — without scheduling operations on stages a product never visits.
• Schantt handles this through per-class routing. Each product class has its own ordered list of required stages. For clear and opaque the list is Cooking, Depositing, Demoulding, Oiling, Packaging — Coating is absent and skipped entirely, producing no operation, no machine assignment, and no Gantt row. For coated the list is Cooking, Depositing, Demoulding, Coating, Packaging — Oiling is skipped. At the split after Demoulding, the transfer times bridge directly: a 15-minute transfer connects Demoulding to Oiling for clear and opaque classes, and a separate 15-minute transfer connects Demoulding to Coating for the coated class, both applied as forward-only delays. At the merge into Packaging, both streams arrive via their own 15-minute transfer (Oiling to Packaging or Coating to Packaging) and the scheduler interleaves them across the three packaging lines according to availability.

4. Balancing parallel machine pools across batch and flow stages.

• The facility has 5 cooking kettles, 6 depositing machines, 6 demould stations, 3 oiling tumblers, 2 coating drums, and 3 packaging lines — each pool with its own processing physics. A kettle cooks a 1,000 kg batch every 45 minutes, a depositing machine runs per-class batches of 160–200 kg with 18–22 minute cycles, and a packaging line processes units at a continuous rate of 1,000–2,000 per hour. An imbalance in one pool creates bottlenecks that ripple downstream: if depositing machines are overloaded, the kettles queue waiting to discharge, and the packaging lines sit idle.
• Schantt treats each stage as a pool of parallel machines. The algorithm explores machine assignments across every pool simultaneously, restricting itself to machines that are capable of the product at that stage. For batch stages (Cooking, Depositing, Demoulding, Oiling, Coating), it computes operation duration as the ceiling of job quantity divided by batch size, multiplied by the cycle duration. For the flow stage (Packaging), it converts the throughput per hour to a per-unit rate and applies it continuously. The solver then assigns each job to the machine combination — one machine per required stage — that minimises total production time, considering calendar availability, upstream-material arrival, and machine-specific changeover times at every stage.

What to model in Schantt

The following five entities form the core of the NutriGummy configuration in Schantt.

Step-by-step setup

1. Create the stages in order. Add six stages in sequence: Cooking, Depositing, Demoulding, Oiling, Coating, and Packaging. Set Cooking, Depositing, Demoulding, Oiling, and Coating to batch production type and Packaging to flow (continuous throughput). On each stage's detail page, configure the transfer times to the next stage — these are the handoff delays between production steps:

• Cooking to Depositing: 15 minutes
• Depositing to Demoulding: per-class drying dwell — 480 minutes (clear, 8 hours), 1,080 minutes (opaque, 18 hours), 720 minutes (coated, 12 hours)
• Demoulding to Oiling: 15 minutes (clear and opaque classes only)
• Demoulding to Coating: 15 minutes (bridging transfer for coated class only)
• Oiling to Packaging: 15 minutes (clear and opaque classes only)
• Coating to Packaging: 15 minutes (coated class only)

2. Add the machines to each stage. Create 25 machines assigned to their respective stages:

• Cooking: Kettle-1 through Kettle-5
• Depositing: Depositor-1 through Depositor-6
• Demoulding: Demould-1 through Demould-6
• Oiling: Oiling-Tumbler-1 through Oiling-Tumbler-3
• Coating: Coating-Drum-1 and Coating-Drum-2
• Packaging: Jar-Line-1, Pouch-Line-2, FlowWrap-Line-3

3. Create the product classes and define per-class routings. Create three product classes — Clear (Gelatin-Based), Opaque (High-Solids), and Coated (Sugar-Polish) — each with the unit set to "unit". On each class's detail page, define the routing:

• Clear: Cooking → Depositing → Demoulding → Oiling → Packaging (skips Coating)
• Opaque: Cooking → Depositing → Demoulding → Oiling → Packaging (skips Coating)
• Coated: Cooking → Depositing → Demoulding → Coating → Packaging (skips Oiling)

No partial transfers are needed since each stage processes the full batch quantity before handoff.

4. Add one representative product per class. Create three products — Clear Gummy (Clear class), Opaque Gummy (Opaque class), Coated Gummy (Coated class) — each inheriting its class's routing and machine parameters. Assign a distinct display colour to each for Gantt readability.

5. Set machine capacity parameters and changeovers. On each machine's detail page, configure the processing parameters and changeover times. Because machines are configured per product class, this step requires the classes from step 3 to exist first.

• Kettles (Cooking): batch size of 1,000 kg, cycle duration of 45 minutes for all three classes. Directional changeovers between every class pair, ranging from 15 to 25 minutes — heavier cleaning when moving from opaque or coated to clear.
• Depositors (Depositing): per-class batch sizes — 160 kg (clear), 200 kg (opaque), 180 kg (coated) — with cycle durations of 18 minutes (clear, coated) and 22 minutes (opaque). Directional changeovers from 45 to 75 minutes, with light-to-dark cleanouts at the higher end.
• Demould stations (Demoulding): match the depositing batch sizes per class, with cycle durations of 10 minutes (clear, coated) and 12 minutes (opaque). Changeovers from 8 to 20 minutes.
• Oiling tumblers (Oiling): batch size matching the incoming batch, cycle duration of 3 minutes per batch. Symmetric 12-minute changeovers between clear and opaque.
• Coating drums (Coating): batch size of 250 kg, cycle duration of 22 minutes for the coated class. Only one class routes here, so no changeovers are required.
• Packaging lines (Packaging): flow stage — set throughput in units per hour. Jar-Line-1 at 1,000 units/hr, Pouch-Line-2 at 1,500 units/hr, FlowWrap-Line-3 at 2,000 units/hr. Symmetric 45-minute changeovers between every class on all three lines.

6. Configure calendars, exceptions, and downtimes (optional). Set up the Standard Two-Shift calendar covering Monday to Friday, 06:00–22:00 as the team default — all machines use this unless overridden. Add two calendar exceptions for non-working days: New Year's Day (1 January) and International Workers' Day (1 May). Optionally add machine downtimes to block capacity for planned outages — a factory-wide year-end shutdown (24–31 December) and a Jar-Line-1 overhaul window (15–17 August). These downtimes are subtracted from available working capacity before scheduling, so no jobs land in the outage window.

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

Common mistakes

1. Modelling drying as a separate stage. Adding a dedicated "Drying" stage creates an unnecessary stage with zero machines and confuses the schedule with an extra operation. The drying dwell is a forward delay — model it as a per-class transfer time from Depositing to Demoulding. Fix: remove the drying stage and enter the dwell durations as transfer times on the Depositing stage detail page, one per class.

2. Using a single blanket changeover time instead of directional per-pair values. A uniform changeover time on depositing machines ignores the fact that switching from opaque to clear takes nearly twice as long as the reverse. The schedule cannot optimise cleanout sequences accurately without directional durations. Fix: enter each directional pair separately — for example, opaque-to-clear at 45 minutes and clear-to-opaque at 75 minutes — so the scheduler can cluster compatible colours.

3. Creating one product class that covers both the clear-or-opaque route and the coated route. Combining all three product types into a single class forces every product through the same routing, losing the ability to skip Oiling for coated gummies or skip Coating for clear ones. Fix: create separate classes for Clear, Opaque, and Coated, each with its own routing that includes only the stages that product type actually visits.

4. Setting packaging lines as batch stages instead of flow stages. Entering a batch size and cycle duration for a packaging line forces it to wait for a full batch to accumulate before processing, when in practice the line runs continuously at a steady rate. Fix: set Packaging to flow production type and enter throughput in units per hour on each line's detail page.

5. Leaving the transfer time from Depositing to Demoulding at the default value. The default transfer time (typically a few minutes) bypasses the multi-hour drying dwell entirely, causing demoulding to start almost immediately after depositing — a physically impossible sequence. Fix: set the per-class transfer times to the realistic drying dwell (8, 12, or 18 hours depending on the class) so the schedule inserts the required hold before demoulding.

What a good schedule looks like

A well-configured Schantt schedule for NutriGummy Co. produces a feasible, shift-respecting plan that accounts for every modelled constraint. The baseline scenario — manual spreadsheet scheduling — typically produces a schedule that is workable only with heavy manual adjustment: changeover downtime consumes 3–6 hours per day, jobs are frequently scheduled outside working hours, and the drying dwell is often treated as a rough buffer rather than an exact delay.

Before (manual spreadsheet):
- 3–6 hours of unaccounted changeover time per day, especially dark-to-light cleanouts on depositing machines, eroding the effective working window
- Jobs manually padded around drying dwell, leading to inconsistent demoulding start times and idle downstream capacity at Oiling, Coating, and Packaging
- Machine assignments determined by planner intuition, often overloading certain depositors or packaging lines while others sit idle
- Weekend gaps and holiday shutdowns applied inconsistently, producing schedules that assign work on non-working days
- Each schedule takes the 2-person planning team significant manual effort to construct and validate before release

After (Schantt Auto mode):
- Changeover time is calculated precisely per machine and product-class pair; the algorithm clusters compatible colours to minimise cleaning downtime across the schedule, recovering productive hours each shift
- Drying dwell is an exact per-class transfer time — demoulding begins the moment the hold elapses, reducing idle wait at downstream stages and keeping the line fed
- All 25 machines are assigned optimally: each job lands on the machine combination that minimises total production time, and parallel pools are balanced to reduce bottlenecks across every stage
- Every operation respects the two-shift calendar, holiday exceptions, and scheduled maintenance windows — no work is assigned outside available hours, so the plan is release-ready without manual correction
- The planning team can review the result in the Gantt view, verify operation timing with calendar overlays, and release to the floor with confidence

The resulting schedule is a shift-aware, constraint-respecting plan that the planning team can review in the Gantt view, verify against the production calendar overlays, and release to the floor — all without the manual rework that spreadsheets require.