Packaged Food Manufacturing Background
Packaged food manufacturing is highly sophisticated with complex multi-line plant layouts, increased order mix, allergen restrictions, and supply chain disruptions. Manually scheduling production using well-worn spreadsheets cannot address such complexity. Eyelit Technologies has the functionality to deal with all the complexity of packaged food manufacturing and will always generate buildable and high-quality schedules that can be executed without further manual introduction. This is illustrated by a use case of scheduling soup filler lines.
Production Scheduling Requirement
In this use case, there are two soup lines with the following scheduling requirements:
- Adhere to a preferred allergen sequence
- Minimize changeovers between item numbers
- Minimize changeovers between allergen families
- Maintain an Allergen-Safe Production Scheduling – Food Manufacturing Case Study stock-on-hand for each item number
- Run the same product concurrently on each of the lines
- Low runners can run on one line while keeping the other lines idle
Batch Sizes and Parallel Batches on Both Lines
In the Grid Browser images below, each block shown represents a batch. As can be seen, items are split equally across both lines, maintaining the 2-batch minimum rule and 2-batch increment rules for additional production.
Changeovers Within Allergen Family
A preferred allergen sequence is defined where allergen codes (E, F, M, S, W) are built up within a run of an allergen family in the following way:
| SEQ. ORDER OF ALLERGENS WITHIN FAMILY | FAMILY 1, GRID VALUE 1 | FAMILY 2, GRID VALUE 3 | FAMILY 3, GRID VALUE 4 | NO ALLERGEN, GRID VALUE 2 |
|---|---|---|---|---|
| 1 | S | M | M | – |
| 2 | S, W | M, W | M, S | – |
| 3 | S, W, M | M, W, E | M, S, W or M, S, F or M, S, SF | – |
| 4 | – | M, W, E, S | – | – |
If this sequence within each family is broken, each break in sequence incurs a changeover time of 2 hours. If the sequence is maintained, the changeover time is reduced to 65–85 minutes, depending on the allergen code.
The allergen sequence shown here demonstrates the buildup of allergen codes within an allergen family in an exhaustive pattern. The key to the right shows individual allergen codes identified by a number 1, 2, 3, or 4. Each time production for a new allergen family occurs, the allergen code associated with number 1 for the new family is produced, and this then results in a sequential buildup within the new allergen family. Items that had allergen codes within the same allergen family were then grouped together to further reduce changeovers:
Order Fulfillment
Each item number is classified as a Made-to-stock product, and order fulfillment is from stock. Finished Goods stock for each item number is modeled as a constraint to be monitored. The grid below shows that each order is fulfilled on time. This means that the schedule ensured that a sufficient stock of each item number is always available when it is time for shipping.
The bar graph below shows an example of the finished goods inventory for one item number (1016219) with a starting inventory of 23 units. The minimum and maximum limits have been obeyed throughout the sequence.
Results
By better balancing the trade-offs involved through patented AI and optimization-based scheduling, Eyelit Technologies can significantly reduce cleaning time and improve delivery times. Adherence to all the requirements can be verified using the many visualizations available in the system, such as the Grid Browser.
The quality of the sequences automatically generated by Eyelit Technologies, as indicated by adherence to each requirement, cannot be realized through spreadsheet tools. Typically, these will require a great deal of manual intervention to fix the many violations of sequencing rules.