Production Scheduling for Snack Food Manufacturing

This guide shows production planners and operations managers how to configure Schantt for snack food manufacturing — a hybrid flowshop where batch cooking and mixing feed continuous frying, baking, seasoning, and packaging lines, all sharing a common fryer bottleneck. You will learn which entities to create, how to set them up in dependency order, and what a well-configured schedule looks like for a multi-product snack plant.

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

Industry context

Snack food production combines several distinct process technologies on a single factory floor. Raw potatoes are washed and sliced; corn is cooked with lime in batch kettles and steeped overnight; flour dough is mixed, extruded, and baked. Fried products — potato chips and tortilla chips — pass through a continuous oil fryer, then receive seasoning in rotating drums, and finally reach vertical form-fill-seal packaging machines. Each of these steps imposes its own pacing, capacity, and changeover characteristics. No two product classes follow the same path through the facility — some skip baking, others skip slicing, and hard pretzels skip both frying and seasoning entirely.

The scheduling challenge in snack manufacturing is one of divergent flows converging on shared equipment. A single continuous fryer running at 1,200 kg/h serves both potato chip and tortilla chip production. A single band oven bakes tortilla sheets and pretzels at different throughput rates. Two seasoning drums share flavour-change work across two product classes, with cleaning times that depend on which seasoning type ran before. Two packaging baggers draw from a common distribution conveyor, and a format change on one bagger forces upstream adjustments. The weekly production plan must sequence all three product classes across these shared resources while respecting a fixed calendar — two shifts Monday through Friday, one shift on Saturday, and scheduled maintenance windows.

Summit Snackcraft runs 22 people at a single 2,600 m² facility, making three product classes — Classic Salted Potato Chips, Restaurant-Style Tortilla Chips, and Classic Hard Pretzels — across seven production stages, scheduled by a production manager supported by two shift supervisors.

Process overview

flowchart LR
    SLICE["SLICE<br/>(Potato slicing)"] --> FRY["FRY<br/>(Continuous fryer)"]
    MASA["MASA<br/>(Corn cooking)"] --> BAKE["BAKE<br/>(Band oven)"]
    DOUGH["DOUGH<br/>(Dough mixing)"] --> BAKE
    BAKE --> FRY
    BAKE --> PACK["PACK<br/>(VFFS bagging)"]
    FRY --> SEASON["SEASON<br/>(Seasoning drums)"]
    SEASON --> PACK

Production flows through seven stages at Summit Snackcraft: raw preparation, cooking and mixing, baking and frying, seasoning, and packaging. Each product class visits a subset of these stages.

Potato chips skip MASA, DOUGH, and BAKE; tortilla chips skip SLICE and DOUGH; pretzels skip SLICE, MASA, FRY, and SEASON.

Potato chips enter at the slicer (SLICE), go directly to the continuous fryer (FRY), then to seasoning (SEASON) and packaging (PACK) — three stages skipped. Tortilla chips start at the masa cooker (MASA), pass through the band oven (BAKE), then join the shared fryer, seasoning, and packaging stages. Pretzels begin at the dough mixer (DOUGH), go straight to the band oven, then cool and pack — skipping four stages.

Scheduling challenges and how Schantt handles them

The weekly production plan at Summit Snackcraft is driven by a demand plan — quantities for each product that must be scheduled within the week. The planner enters demand for each product as a total quantity, and Schantt schedules the required production runs forward from a chosen start date. If your schedule is driven by order backlog or a rolling sales forecast instead, the same workflow applies: convert your upcoming orders into demand entries for each product in the system.

Schantt optimises total production time — the duration from the first operation to the last. It explores alternative machine assignments and job sequences to find the combination that finishes all work in the shortest time. This guide assumes a weekly scheduling horizon, which matches the typical planning cycle at a mid-market snack plant.

Two scheduling modes are available. Auto mode reorders jobs automatically to find the sequence that minimises total production time. Semi-Auto mode preserves the planner's job order — for example, running all dairy-seasoned products first to minimise allergen cleaning — but still optimises machine assignments within that fixed sequence.

What Schantt handles well

• Multi-product hybrid flowshop with mixed batch-and-flow pipelines — models both continuous fryers and ovens (flow) and batch cookers and mixers (batch) in a single sequential route, with automatic wait-material pauses when a downstream stage outruns its supply.

• Sequence-dependent flavour changeovers — a directional per-machine changeover matrix groups same-flavour-family runs and minimises changeover time; Auto mode reorders jobs, Semi-Auto mode optimises within the planner's fixed order.

• Parallel-machine allocation for fryer-to-packaging distribution — explores which machine each job lands on across multi-machine stages, choosing the combination that minimises total production time.

• Per-class routing with stage skipping — each product class visits only its required stages; skipped stages produce no operations, and bridging transfer times connect across the gap.

• Calendar-aware working capacity and seasonal shifts — shift patterns, holiday exceptions, overtime dates, and scheduled maintenance windows all affect timing automatically and render as coloured overlays on the Gantt.

• Bottleneck emergence from flow-stage throughput modeling — the bottleneck emerges naturally from throughput values; no manual designation is needed.

How Schantt handles each challenge

1. Fryer bottleneck constrains total plant output.

• The single continuous fryer limits throughput to 1,200 kg/h — roughly 19,000 kg per two-shift day on potato chips. Any downtime, from product changes or scheduled maintenance, is permanently lost because the fryer cannot be run faster to catch up. Typical downtime losses of 2–3 hours per week reduce effective output by 5–7%. The fryer's throughput is modelled as a per-class rate on the machine, with 1,200 kg/h for both potato chips and tortilla chips. Because the fryer is the only machine on the FRY stage and both product classes pass through it, the bottleneck emerges naturally — downstream seasoning and packaging stages that could run faster are paced by the fryer's supply rate, and wait-material pauses appear automatically on the Gantt where downstream capacity exceeds what arrives.
• Scheduled events such as the quarterly oil change (6–8 hours on July 15) are entered as planned downtime that blocks fryer availability. The schedule routes production around the blocked window. Changeover time on the fryer (15 minutes between potato chips and tortilla chips in either direction) is modelled as a directional setup on the machine, so the scheduler accounts for the pause whenever a recipe switch occurs.

2. Frequent flavour changeovers consume a large share of seasoning capacity.

• Four to six flavour transitions per shift on two seasoning drums, each taking 15–30 minutes within the same allergen group, add up to 1.5–3 hours of changeover time per shift — roughly 15–25% of seasoning-stage capacity. On a high-mix day the seasoning drums spend more time being cleaned and set up than they do applying seasoning. The reality is that each transition is direction-dependent but the durations are predictable once the product sequence is known. The planner must decide which product class runs next on which drum, and each wrong choice adds wasted setup time.
• The directional per-machine changeover matrix captures each flavour transition's time penalty on each drum. In Auto mode, the scheduler groups same-product-class runs together, reducing the total number of transitions. In Semi-Auto mode, the planner locks in a desired sequence — for example, running all potato chip flavours as a single block — and Schantt still optimises which drum handles each job within that fixed order.

3. Tight transfer timing between fryer and seasoning.

• Fried products must reach the seasoning drum within 60 seconds of leaving the fryer. The conveyor between them takes 30 seconds minimum, but any seasoning drum stoppage backs the fryer up within minutes. A 10-minute unplanned stop can compromise roughly 200 kg of hot product. The conveyor handoff is modelled as a transfer time of 0.5 minutes between the FRY and SEASON stages, so the schedule chains each fryer run to a seasoning-drum start at that minimum delay. The transfer time captures the minimum conveyor dwell but not the maximum 60-second window — that constraint is a planner Gantt check.
• The planner inspects each fryer block on the Gantt to confirm that a seasoning drum is free within the viable window. If a fryer run's seasoning start is too late, the planner re-sequences the affected jobs in Semi-Auto mode — locking in a safe order while keeping machine assignments optimised.

4. Allergen segregation creates asymmetric cleaning times across seasoning drums.

• One drum is dedicated to non-allergen dry seasonings; the other handles dairy-based powders. Switching from dairy to non-allergen requires 45–60 minutes of cleaning, while the reverse takes only 15–20 minutes. The scheduling reality is that a non-allergen-to-dairy transition is short (15–20 minutes) but a dairy-to-non-allergen transition is long (45–60 minutes). The planner needs the schedule to apply the correct time penalty based on the transition direction.
• Directional changeover times on each drum capture the duration difference between allergen-group transitions. Schantt applies the correct time penalty based on direction of each change — a dairy-to-non-allergen transition on drum-02 generates the longer duration automatically, and a non-allergen-to-dairy transition on the same drum generates the shorter one. The planner sets the production order — for example, running all dairy-seasoned products consecutively — via Semi-Auto mode's fixed sequence or verifies the grouping on the Gantt.

5. Seasonal demand peaks require periodic capacity adjustments.

• Summer barbecue season and year-end holidays drive demand 1.5–2 times baseline for six to eight weeks total, requiring additional shifts beyond the standard two-shift Monday–Saturday pattern. Capacity-expansion decisions — when to add a third shift and by how much — are made outside Schantt based on sales forecasts and commercial targets. The planner knows the seasonal dates and decides the adjusted working hours manually.
• Calendar mechanics let the planner express that decision directly: edit the active shift pattern to add a third shift for the relevant date range, and enter calendar exceptions for specific overtime dates. The schedule recomputes against the adjusted working hours, with added capacity visible as extended working-day overlays on the Gantt. The capacity-adjustment amounts and timing remain planner input — Schantt applies whatever working hours the planner defines.

What to model in Schantt

One team owns one schedule for Summit Snackcraft's single facility. The configuration is built from five entity types:

Step-by-step setup

1. Create the stages in production order. Define the seven stages in sequence — SLICE, MASA, DOUGH, BAKE, FRY, SEASON, PACK — setting each to flow or batch according to its production type. Then, on each stage's detail page, enter the transfer times that connect consecutive stages used by any routing. With three divergent routings, seven transfer-time entries are needed:

• Forward flows: SLICE → FRY (1 min), FRY → SEASON (0.5 min), SEASON → PACK (4 min)
• Tortilla path: MASA → BAKE (5 min), BAKE → FRY (2 min)
• Pretzel path: DOUGH → BAKE (3 min)
• Skip bridge: BAKE → PACK (6 min) — connects baking directly to packaging for pretzels

2. Add the machines to each stage. Assign the nine machines to their stages: one slicer on SLICE, one masa cooker on MASA, one dough mixer on DOUGH, one band oven on BAKE, one continuous fryer on FRY, two seasoning drums on SEASON, and two VFFS baggers on PACK.

3. Create the product classes and define per-class routings. Create three product classes — Classic Salted Potato Chips, Restaurant-Style Tortilla Chips, and Classic Hard Pretzels. On each class's detail page, select only the stages its routing visits and leave the skipped stages unselected. Partial-transfer toggles should be left off for all three classes.

4. Add one product per class. Create one representative product for each class — for example, a 150 g bag of salted potato chips assigned to the Classic Salted Potato Chips product class, a 300 g bag of tortilla chips under its class, and a 200 g bag of hard pretzels under the Classic Hard Pretzels class. Assign each product a Gantt display colour so jobs are easily distinguishable on the schedule. One product per class is sufficient for a planning-level model — additional pack sizes or flavour variants within a class inherit the same routing and machine parameters without needing separate product entries.

5. Set machine capacity parameters and changeovers. On each machine's detail page, configure the per-class parameters that drive timing. Capacity settings are entered per product class because each class may run at a different rate on the same machine:

• Throughput (flow stages):
• SLICE — slicer-01: 3,000 kg/h (potato chips)
• FRY — fryer-01: 1,200 kg/h (potato chips and tortilla chips)
• SEASON — drum-01 and drum-02: 1,200 kg/h each (both product classes)
• BAKE — oven-01: 400 kg/h (tortilla chips), 250 kg/h (pretzels)

• PACK — bagger-01: 600 kg/h (potato chips); bagger-02: 500 kg/h (tortilla chips and pretzels)

• Batch cycle and size (batch stages):

• MASA — cooker-01: 740 min cycle, 500 kg batch

• DOUGH — mixer-01: 12 min cycle, 200 kg batch

• Directional changeovers — enter per-pair setup times on the four machines shared by multiple product classes:

• Oven: 20 min between tortilla chips and pretzels (both directions)
• Fryer: 15 min between potato chips and tortilla chips (both directions)
• Seasoning drum-01: 10 min between potato chips and tortilla chips (both directions)
• Bagger-02: 25 min between tortilla chips and pretzels (both directions)

6. Configure the calendar, exceptions, and downtimes. Create a single default calendar with Monday–Friday two shifts (06:00–22:00) and Saturday one shift (06:00–14:00). Sunday is non-working. Then add three calendar exceptions for non-working days (January 1, May 1, December 31) and two machine downtimes: the quarterly fryer oil change (July 15, 06:00–14:00) and the year-end facility shutdown (December 31, 06:00–18:00).

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 instead of per-pair directional values. Applying one changeover duration to all transitions ignores the real difference between a 15-minute fryer change and a 25-minute bagger format change. The scheduler cannot apply the right time penalty per transition, producing unrealistic gaps. Fix: Enter directional changeover times for every pair of product classes that share a machine, using the durations that match each direction.

2. Grouping all products under one product class with a single routing. Potato chips, tortilla chips, and pretzels follow completely different paths and skip different stages. A single routing would force every product through every stage, generating empty operations on machines that product never visits. Fix: Create a separate product class for each distinct routing variant, with its own stage list and skipped-stage entries.

3. Forgetting skip-bridge transfer times between disconnected stages. When a product class skips several interior stages — for example, pretzels skipping from DOUGH directly to BAKE — the transfer time for the skipped span must still be entered as a bridge entry. Without it, the handoff delay is zero and the schedule shows the next operation starting immediately, which does not reflect the real conveyor travel time. Fix: For every pair of consecutive routed stages that are not adjacent in the full stage list, add a bridging transfer-time entry between them.

4. Setting a batch stage's throughput instead of its batch size and cycle time. Batch cookers and mixers do not run at a continuous rate — they process discrete loads with a fixed cycle duration. Entering a throughput value for MASA or DOUGH would produce incorrect timing because the scheduler would treat them as flow stages. Fix: On batch stages, set the batch size (kg per load) and the cycle duration (minutes per batch), leaving throughput blank.

5. Not accounting for calendar exceptions when scheduling across holidays. A schedule that ignores January 1, May 1, and the year-end shutdown will place operations on non-working days, producing plans that cannot be executed. Fix: Enter all calendar exceptions before generating the schedule, and add scheduled maintenance downtimes so blocked availability is visible on the Gantt.

What a good schedule looks like

A well-configured schedule in Schantt translates Summit Snackcraft's weekly demand plan into a timed, machine-by-machine production sequence where every operation has a start time and a duration that respects the plant's real constraints.

Before Summit Snackcraft adopted structured scheduling, the weekly plan was a manual sequence on a whiteboard. The fryer bottleneck was not explicitly accounted for, seasoning changeovers piled up unpredictably on high-mix days, and packaging format changes forced last-minute re-planning. Symptoms included:

• Total changeover time consuming as much as 25% of available seasoning capacity
• Unplanned fryer-to-seasoning handoff conflicts requiring mid-shift corrections
• Packaging backups when a bagger format change was scheduled without checking whether the remaining bagger could handle the combined output

Before (manual whiteboard): The planner estimated run times by experience and wrote start times on a whiteboard each morning. Every product change meant checking each shared machine's status manually. A fryer oil change or holiday shutdown required re-drawing large sections of the schedule by hand, and the fifteen-minute fryer changeover between potato chips and tortilla chips was often forgotten until the shift supervisor flagged it mid-production.

After (Schantt Auto mode): With the seven stages, nine machines, three product classes, and directional changeovers configured, the scheduler produces a Gantt showing every operation in timed sequence:

• The schedule groups same-product-class runs together, reducing the number of changeovers on the fryer, oven, and seasoning drums. Where manual scheduling averaged 4–6 flavour changes per shift, the optimised sequence consolidates them into fewer transitions.
• Fryer throughput at 1,200 kg/h emerges as the natural pacing rate for all fried products. Downstream seasoning and packaging operations that could run faster show wait-material pauses where they have caught up and are waiting for the next batch to arrive from the fryer.
• Each product class appears on only its own stages. The Gantt has no rows for skipped stages — potato chips show no baking operations, pretzels show no frying operations — making the schedule easier to read at a glance.
• The quarterly oil change on July 15 and the year-end shutdown on December 31 are visible as blocked time on the affected machines. The scheduler routes production around these windows automatically, without manual adjustment.
• The planner inspects the fryer-to-seasoning handoff timing on the Gantt to confirm each fryer block's seasoning drum is available within the 60-second window. Where the handoff is tight, the planner switches to Semi-Auto mode to lock in a specific sequence while still getting optimised machine assignments.