Multi-Distribution Center (DC) Planning

Many logistics operations involve more than one distribution center or depot. Planners for each DC often work independently, managing their own fleets, orders, and operational constraints. SWAT API provides a powerful way to manage multi-DC operations within a single project, ensuring both separation of concerns and centralized oversight.

The Challenge of Multi-DC Planning

Traditionally, managing multiple DCs might require separate system accounts or complex data filtering to prevent planners from one DC from accidentally modifying data for another. This can lead to inefficiencies and potential errors.

Simulation Templates per DC

The recommended approach for multi-DC planning in SWAT is to use a separate simulation template for each distribution center. This allows each DC to have its own isolated planning environment while still operating under a single, unified project.

How It Works

1. Single Project: Your entire operation, across all DCs, is contained within a single SWAT project.
2. One Template per DC: For each distribution center, you create a dedicated simulation_template. For example, if you have a "North DC" and a "South DC", you would create two templates: template_dc_north and template_dc_south.
3. Customized Settings: Each simulation template contains the specific configuration for its corresponding DC:
   • Fleet: Only the vehicles that operate out of that DC.
   • Constraints: Operational rules, time windows, and other constraints specific to that DC's region or operational model.
   • Optimization Goals: The cost functions and optimization priorities for that DC.
4. Daily Operations: When a planner for a specific DC starts their work for the day, they create a new simulation by cloning it from their DC's dedicated template. This new simulation inherits all the correct settings.
5. Separation of Planning: The planner then uploads their daily orders into this new simulation and runs the optimization. Because each simulation is self-contained with the vehicles and settings for a specific DC, the planning process is neatly separated. This prevents resources from one DC (e.g., a truck from the North DC) from being accidentally used in a plan for another DC. While a planner with appropriate permissions might be able to view simulations for multiple DCs in the same UI, the optimization for each DC is run independently, ensuring operational integrity.

Benefits of This Approach

• Centralized Management, Segregated Planning: You can manage your entire operation through a single API project, but daily planning is neatly segregated by DC.
• Independent Configurations: Each DC can have vastly different operational settings without any risk of conflict. For example, the North DC might use electric vehicles with different range constraints, while the South DC uses diesel trucks.
• Scalability: Adding a new DC is as simple as creating a new simulation template for it.
• Consistency: All planners use the same API workflows, even though their underlying data and settings are different. This simplifies integration and training.

Example Scenario

Imagine a company with two warehouses, "Warehouse A" and "Warehouse B".

1. Setup:

   • A single project is created in SWAT.
   • Two simulation templates are created:
     • template_warehouse_a: Contains the 10 trucks that operate from Warehouse A and the rule to minimize mileage.
     • template_warehouse_b: Contains the 15 vans that operate from Warehouse B and the rule to prioritize on-time deliveries.

2. Planning:

   • Planner A needs to plan routes for Warehouse A. They make an API call to create a new simulation for the day, cloning it from template_warehouse_a. They then upload the day's orders for Warehouse A into this new simulation and run the optimization. The result will only use the 10 trucks from Warehouse A.
   • Planner B does the same, but they clone from template_warehouse_b. Their optimization will only use the 15 vans from Warehouse B and will focus on meeting time windows.

Both planners can work simultaneously without any interference, using the same API endpoints but pointing to their respective simulation templates. This provides a clean, scalable, and efficient way to manage multi-DC logistics operations.