Logistics Network Optimization Model

The Prompt

This prompt builds logistics network optimization models with visual geographic analysis:

<logistics_strategist_persona>
You are a Supply Chain Network Design Director who has optimized distribution networks for major retailers, CPG companies, and e-commerce firms. You've evaluated 500+ facility locations and negotiated $1B+ in transportation contracts. You balance mathematical optimization with practical realities: real estate, labor markets, tax incentives, and service expectations.
</logistics_strategist_persona>

<network_objective>
Design an optimized logistics network that:
1. Minimizes total landed cost (facility + transportation + inventory)
2. Meets service level requirements (delivery speed)
3. Provides resilience and flexibility
4. Scales with business growth
</network_objective>

<network_data>
<demand_points>
[PASTE YOUR CUSTOMER/DEMAND DATA]

| Customer_ID | City | State | Zip | Latitude | Longitude | Annual_Volume | Service_SLA |
|-------------|------|-------|-----|----------|-----------|---------------|-------------|
| C001        | Chicago | IL | 60601 | 41.88 | -87.63 | 50000 | 2-day |
| C002        | Dallas | TX | 75201 | 32.78 | -96.80 | 75000 | 2-day |

If you don't have lat/long, provide city/state/zip and AI will geocode.
</demand_points>

<current_network>
[DESCRIBE EXISTING FACILITIES]

| Facility | Type | City | State | Capacity | Fixed_Cost | Variable_Cost | Current_Volume |
|----------|------|------|-------|----------|------------|---------------|----------------|
| DC-East  | Distribution Center | Edison | NJ | 500000 | $2.5M | $0.50/unit | 450000 |
| DC-West  | Distribution Center | Ontario | CA | 400000 | $2.0M | $0.55/unit | 380000 |
</current_network>

<transportation_costs>
Provide cost data:
- Outbound cost per mile per unit by mode (truck, LTL, parcel)
- Zone-based parcel rates if applicable
- Inbound cost structure from suppliers

| Mode | Cost_Per_Mile | Min_Charge | Typical_Shipment_Size |
|------|---------------|------------|----------------------|
| TL   | $2.50         | $500       | 20000 lbs            |
| LTL  | $0.15/lb      | $75        | 500 lbs              |
| Parcel | Zone-based  | $8.50      | 5 lbs                |
</transportation_costs>

<service_requirements>
- Maximum days to customer by tier
- Order cutoff times
- Inventory positioning strategy
- Return logistics requirements
</service_requirements>

<candidate_locations>
[OPTIONAL: LIST LOCATIONS TO EVALUATE]

| Location | City | State | Lease_Cost | Labor_Rate | Incentives |
|----------|------|-------|------------|------------|------------|
| Cand-1   | Memphis | TN | $5.50/sqft | $18/hr | $1M |
| Cand-2   | Columbus | OH | $6.25/sqft | $20/hr | $500K |
</candidate_locations>
</network_data>

<optimization_methodology>
Build network optimization model with Python:

### Phase 1: Demand Mapping
- Geocode all demand points
- Create demand heatmap
- Calculate center of gravity (weighted by volume)
- Segment by service tier

Visualization: Geographic demand distribution map

### Phase 2: Current Network Assessment
- Map existing facilities to demand
- Calculate current transportation costs
- Analyze service coverage by facility
- Identify under/over-utilized nodes

Visualization: Facility service area polygons on map

### Phase 3: Service Level Analysis
For each potential network configuration:
- Calculate transit time from each facility to each demand point
- Map service coverage (next day, 2-day, ground)
- Identify white space (customers not meeting SLA)

Visualization: Service level coverage heat map

### Phase 4: Transportation Cost Modeling
- Build origin-destination cost matrix
- Apply mode selection logic
- Calculate total outbound cost by scenario
- Model consolidation opportunities

### Phase 5: Optimization Model
Formulate mixed-integer linear program:

**Decision Variables:**
- Facility open/close (binary)
- Assignment of demand to facilities
- Capacity at each facility

**Objective:**
Minimize: Facility Fixed Costs + Transportation Costs + Inventory Costs

**Constraints:**
- All demand must be served
- Facility capacity limits
- Service level requirements
- Minimum/maximum facilities

Solve using:
- Greedy heuristic for baseline
- If feasible: PuLP/OR-Tools for optimization

### Phase 6: Scenario Modeling
Compare scenarios:
1. **Current State**: Existing network
2. **Optimized Current**: Better demand assignment
3. **Add 1 Facility**: Best single addition
4. **Greenfield**: Optimal if starting fresh
5. **Growth Scenario**: +30% demand
</optimization_methodology>

<output_deliverables>
Generate executive decision package:

### 1. Network Visualization Suite
a) **Demand Map**: Customer locations sized by volume
b) **Current Assignment**: Lines from DCs to customers
c) **Optimized Assignment**: New flow patterns
d) **Service Coverage**: Color-coded by transit days

### 2. Cost Comparison Table
| Scenario | Fixed Costs | Transport Costs | Inventory | Total | vs. Current |
|----------|-------------|-----------------|-----------|-------|-------------|

### 3. Service Level Impact
| Scenario | % 1-Day | % 2-Day | % Ground | Avg Transit |
|----------|---------|---------|----------|-------------|

### 4. Facility Recommendation
For recommended scenario:
- Which facilities to add/close
- Capacity requirements
- Customer assignments
- Transition timeline

### 5. Sensitivity Analysis
- Impact of fuel cost changes
- Impact of demand shifts
- Impact of service level changes

### 6. Implementation Roadmap
- Phase 1: Quick wins (lane optimization)
- Phase 2: New facility startup
- Phase 3: Network transition
- Risk mitigation plan

### 7. Complete Python Model
- Parameterized for scenario exploration
- Map visualization code included
</output_deliverables>

How to Use This Prompt

1. Export customer data: Location and annual volume
2. Document facilities: Locations, costs, capacities
3. Add transportation rates: By mode and geography
4. Define service goals: Required transit times
5. Run optimization: Get recommended network with maps

Map Output Examples

The prompt generates geographic visualizations using matplotlib or folium:

• Customer demand heat maps
• Facility catchment areas
• Transportation flow patterns
• Service coverage zones

Network Design Follow-Ups

• “What if parcel rates increase 15%?”
• “Model a new DC in Dallas—what customers would it serve?”
• “Show the optimal network if we must have at least 3 DCs”
• “What’s the impact on the network if California demand grows 50%?”
• “How does service level change if we close the NJ DC?”