-DeliverX-Optimizer

Real-time delivery route optimization dashboard powered by Genetic Algorithms.

Watch AI untangle messy delivery routes into optimal paths on a live Google Maps interface — built with Streamlit, Plotly, and NumPy.

---

Overview

SkyRoute-AI is an interactive logistics dashboard that solves the Travelling Salesman Problem (TSP) using a Genetic Algorithm. It generates delivery stops across real-world cities, optimizes the route in real time, and visualizes every generation of the evolution on a live map.

Key Features

• Live Map Visualization — Routes evolve on Google Maps / CartoDB / OpenStreetMap tiles in real time
• 6 Map Styles — Google Maps (Light), Google Satellite, Google Terrain, CartoDB Light, CartoDB Dark, OpenStreetMap
• 9 City Presets — New York, San Francisco, London, Tokyo, Mumbai, Dubai, Sydney, Paris, Singapore
• Optimization Intelligence Panel — Circular gauge, convergence sparkline, distance reduction bar, and evolution progress
• Genetic Algorithm Engine — Tournament Selection, Ordered Crossover (OX1), Swap Mutation, Elitism
• Haversine Distance — Geographically accurate route distances in kilometres
• JSON Export — Download the optimized route with coordinates for integration with driver apps
• Fully Configurable — Tune population size, mutation rate, generations, tournament size, and more

---

Demo

Sidebar Controls          Live Map (Google Maps)        Optimization Intelligence
┌──────────────────┐     ┌──────────────────────┐     ┌──────────────────────┐
│ City: Mumbai     │     │                      │     │   ┌──────┐           │
│ Stops: 30        │     │   Routes untangle    │     │   │58.6% │  Initial  │
│ Population: 100  │     │   in real time on     │     │   │ OPTIM│  130.7 km│
│ Mutation: 0.020  │     │   actual city maps    │     │   └──────┘  Current │
│ Generations: 500 │     │                      │     │              54.2 km │
│                  │     │  ★ DEPOT              │     │ ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔  │
│ [Optimize Route] │     │  ● Stop 1 → Stop 2   │     │ Convergence Curve   │
└──────────────────┘     └──────────────────────┘     │ ╲__________         │
                                                      └──────────────────────┘

---

Quick Start

1. Clone the repository

git clone https://github.com/yourusername/SkyRoute-AI.git
cd SkyRoute-AI

2. Install dependencies

pip install -r requirements.txt

3. Run the app

streamlit run app.py

The dashboard opens at http://localhost:8501.

---

How It Works

The Genetic Algorithm

The RouteOptimizer class implements a classic GA to solve the TSP:

Step	Method	Description
Encoding	Permutation	Each individual is a permutation of [0, 1, ..., n-1] representing visit order
Fitness	1 / distance	Lower total route distance = higher fitness
Selection	Tournament	Pick k random individuals, select the fittest
Crossover	Ordered (OX1)	Copy a segment from parent 1, fill remainder from parent 2 preserving order
Mutation	Swap	With probability p, swap two random stops
Elitism	Top-k carry	Best individuals survive unchanged into the next generation
Distance	Haversine	Great-circle distance between lat/lon coordinates in km

Architecture

app.py (single file)
├── RouteOptimizer        Pure NumPy GA engine (no Streamlit dependency)
├── build_map_figure()    Plotly Scattermapbox with configurable tile layers
├── build_progress_html() Self-contained HTML/SVG stats panel
├── build_route_json()    JSON export for driver app integration
└── Streamlit UI          Sidebar controls, metrics, live map, download button

The RouteOptimizer class has zero knowledge of Streamlit — it takes NumPy arrays in, returns dicts out. This means it can be imported and used in any Python project independently.

---

Configuration

Sidebar Controls

Parameter	Range	Default	Effect
City	9 presets	New York	Determines geographic center for delivery stops
Delivery stops	5 – 80	20	Number of locations to visit
Random seed	0 – 99999	42	Reproducible point generation
Population size	20 – 500	100	Number of route candidates per generation
Mutation rate	0.001 – 0.300	0.020	Probability of random swap per child
Generations	50 – 2000	500	Total evolution cycles
Tournament size	2 – 10	5	Selection pressure (higher = more greedy)
Elite count	1 – 10	2	Guaranteed survivors per generation
Map refresh	1 – 50	10	Update visuals every N generations

Map Styles

Style	Tiles	Best For
Google Maps (Light)	Google road map	General use, familiar look
Google Satellite	Google aerial imagery	Geographic context
Google Terrain	Google topographic	Elevation-aware planning
CartoDB Light	Carto Positron	Clean, minimal presentation
CartoDB Dark	Carto Dark Matter	Dark-themed dashboards
OpenStreetMap	OSM standard	Detailed street-level view

---

JSON Export Format

After optimization, click "Download Optimized Route" to get a JSON file:

{
  "city": "Mumbai, India",
  "total_distance_km": 54.1923,
  "num_stops": 20,
  "route": [
    {
      "visit_order": 0,
      "stop_id": 7,
      "latitude": 19.0834,
      "longitude": 72.8891
    },
    {
      "visit_order": 1,
      "stop_id": 3,
      "latitude": 19.0612,
      "longitude": 72.8543
    }
  ]
}

---

Tech Stack

Component	Technology	Purpose
UI Framework	Streamlit	Interactive dashboard with sidebar controls
Mapping	Plotly Scattermapbox	Real-time route visualization on map tiles
Optimization	NumPy + custom GA	Genetic Algorithm for TSP solving
Distance	Haversine formula	Geographically accurate route distances
Stats Panel	HTML/SVG (inline)	Circular gauge, sparkline, progress bars
Export	JSON (stdlib)	Route data for downstream systems

---

Project Structure

SkyRoute-AI/
├── app.py               # Complete application (GA engine + UI)
├── requirements.txt     # Python dependencies
└── README.md            # This file

---

Performance

Scenario	Stops	Population	Generations	Approx. Time
Quick demo	10	50	200	< 2 seconds
Standard	30	100	500	~ 5 seconds
Heavy	50	200	1000	~ 20 seconds
Maximum	80	500	2000	~ 90 seconds

---

License

This project is licensed under the MIT License. See LICENSE for details.

---

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -m "Add your feature")
4. Push to the branch (git push origin feature/your-feature)
5. Open a Pull Request

---

Built with Streamlit, Plotly, and NumPy
SkyRoute-AI — Smarter routes, faster deliveries.