Example Problem: FreshHarvest Food Delivery Network

To illustrate how cuOpt solves real-world vehicle routing problems (VRP), consider a fictional company Fresh Harvest, a regional food distribution company responsible for delivering perishable and non-perishable goods across a mixed network of retail locations.

Download this example and see it solve in 5 minutes.

FreshHarvest Routing Example.xlsx

Network Overview

The Fresh Harvest distribution network includes:

• 3 warehouses (depots)
  • All vehicles start and end their routes at one of these locations
• 200 delivery sites, consisting of:
  • 50 grocery stores
  • 50 restaurants
  • 100 convenience stores

Each site requires a combination of product types and has its own delivery constraints.

Product Types

Each delivery site demands a mix of the following item categories:

1. Bakery
2. Beverage
3. Dairy
4. Deli
5. Dry goods
6. Frozen
7. Household
8. Poultry
9. Produce
10. Seafood

Each item type contributes to vehicle capacity usage (pallets and weight) and service time, based on unloading requirements

Fleet Composition

Fresh Harvest operates a heterogeneous fleet of 100 vehicles:

In addition to fixed costs, each vehicle type incurs a per-mile cost, with refrigerated vehicles being the highest cost and standard vehicles being the lowest cost.

Time Constraints

• Planning horizon: 720 minutes (12 hours)
• Shift duration: Each vehicle may operate up to 480 minutes (8 hours) within this window
• Each vehicle must take a 30-minute break starting between 180 and 270 minutes into its shift.

Site Constraints

Each delivery location has a defined time window when deliveries are accepted. This introduces scheduling complexity, as routes must align with site availability.

Task-Vehicle Compatibility

Certain deliveries require specific vehicle types:

Refrigerated vehicle-only deliveries:

• Frozen
• Poultry
• Seafood
• Dairy

Urban delivery vehicles are required for sites on these streets:

• Dogwood St
• Maple St
• Poplar St
• Sycamore St

All other deliveries may be handled by compatible vehicle types, subject to capacity and cost considerations.

Objective

The goal is to determine optimal routes that:

1. Minimize the number of vehicles used
2. Minimize the total cost (fixed + distance-based)

The solution also

• Respects all vehicle capacities, both measured by number of pallets and by weight
• Satisfies all delivery time windows
• Enforces proper breaks
• Ensure proper vehicle–task compatibility

This example highlights how cuOpt handles real-world routing complexity, including heterogeneous fleets, time windows, labor rules, and product-specific constraints—all within a single optimization model.