asrulharunismail/FTZ_MULTI-MODAL_MEDICAL_LOGISTICS

🚛 Multi-Modal Vehicle Routing Optimization System

An advanced optimization system for last-mile delivery using bikes, vans, and drones with emissions-focused routing.

🌟 Features

• —Multi-Modal Routing: Optimizes delivery routes using bikes, electric vans, and drones
• —Emissions Focus: 92% weight on emissions reduction for sustainable logistics
• —City Selection: Pre-configured datasets for Southampton and Portsmouth
• —Pair Constraints: Ensures drone pairs operate within time limits
• —Interactive Visualization: Folium maps showing all optimized routes
• —Complete Pipeline: P-Hub Allocation → TSP → Vehicle Routing with Bees Algorithm

🏙️ Available Cities

Southampton (Matrices2PY.xlsx)

• —78 nodes dataset
• —Full network coverage
• —Multiple depot configuration

Portsmouth (Matrices1PY.xlsx)

• —Alternative city configuration
• —Optimized for different network topology

📊 Optimization Pipeline

1. 1.Phase 1: P-Hub Allocation & Bike Routes
2. 2.Allocates customers to hubs
3. 3.Generates bike delivery routes

1. 1.Phase 1.5: TSP Input Preparation
2. 2.Filters zero-demand nodes
3. 3.Prepares matrices for TSP

1. 1.Phase 2: TSP Optimization
2. 2.Domino algorithm for tour construction
3. 3.Optimizes customer visit sequence

1. 1.Phase 2.5: Matrix Processing
2. 2.Depot relabeling (0→101, 1→102)
3. 3.Final matrix preparation

1. 1.Phase 3: Vehicle Routing with Bees Algorithm
2. 2.Greedy segment construction
3. 3.Bees Algorithm optimization
4. 4.Pair constraint enforcement

1. 1.Phase 4: Cost Breakdown Analysis
2. 2.Detailed cost per segment
3. 3.Emission calculations
4. 4.Penalty analysis

1. 1.Phase 5: Schedule Generation
2. 2.Vehicle scheduling with time windows
3. 3.Route consolidation
4. 4.Fleet composition analysis

⚙️ Key Parameters

Schedule Parameters

• —Start Time: When all vehicles begin operations (default: 09:00)
• —Gap Time: Minutes between routes on same vehicle (default: 5 min)
• —Max Operation Time: Maximum vehicle operation duration (default: 90 min)

Vehicle Constraints

• —Max Drones: Maximum number of drone vehicles (default: 6)
• —Max Pair Duration: Combined time limit for drone pairs (default: 90 min)
• —Service Time: Service time per customer (default: 2.5 min)

Objective Function Weights

• —Fixed Cost: Weight for fixed vehicle costs (default: 0.01)
• —Variable Cost: Weight for distance-based costs (default: 0.01)
• —Labour Cost: Weight for time-based costs (default: 0.01)
• —Emissions: Weight for emission costs (default: 0.92) ⭐

Optimization Parameters

• —Number of Bees: Population size (default: 10)
• —Max Iterations: Optimization iterations (default: 1000)

📈 Outputs

Data Files

• —Cost_Breakdown.xlsx: Detailed cost analysis per route
• —Vehicle_Schedule.xlsx: Complete schedule with departure times
• —Vehicle_Statistics.xlsx: Fleet composition summary
• —Best_BA_Solution.xlsx: Optimal solution details

Visualizations

• —Vehicle_Ratio_Chart.png: Fleet composition bar chart
• —Route_Map.html: Interactive route visualization

Analysis

• —Cost breakdown by vehicle type
• —Emission analysis (kg CO₂)
• —Pair constraint compliance
• —Schedule feasibility report

🚀 How to Use

1. 1.Select City: Choose Southampton or Portsmouth
2. 2.Configure Parameters: Adjust optimization settings
3. 3.Run Optimization: Click "Run Optimization" button
4. 4.View Results: Check Results and Visualizations tabs
5. 5.Download Files: Get all output files from Download tab

🔬 Technical Details

Vehicle Types

• —Bike: Capacity 0kg, Cost £3.15 + £1.15/mile
• —Van-D1: Capacity 300kg, 90 min max time, Depot 1
• —Van-D2: Capacity 300kg, 53 min max time, Depot 2
• —Drone: Capacity 75kg, 55 min max time, Zero emissions

Algorithms

• —P-Hub Allocation: Hub location and customer assignment
• —TSP Domino: Tour construction for customer sequence
• —Bees Algorithm: Metaheuristic for route optimization
• —Foraging: Neighborhood search operators

Constraints

• —Vehicle capacity limits
• —Time window constraints
• —Drone pair duration limits (90 min combined)
• —Single customer per drone route

📚 Research Context

This system is designed for last-mile delivery electrification research, focusing on:

• —Carbon emission reduction
• —Multi-modal transportation optimization
• —Sustainable urban logistics
• —Digital twin technology for transport decarbonization

🛠️ Technology Stack

• —Python 3.8+
• —Gradio: Web interface
• —Pandas: Data processing
• —NumPy: Numerical computations
• —Folium: Interactive mapping
• —Matplotlib: Visualization
• —Gurobi (optional): Advanced optimization

👨‍🔬 Author

Developed at Heriot-Watt University for Transit Research and Transport Decarbonization.

📄 License

MIT License

🤝 Acknowledgments

• —CLD1 and CLD3 Project Partners
• —Industry collaborators (DHL)
• —Heriot-Watt University

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference