🚖 SmartCommute: Live Multimodal Travel & Fare Optimizer (Alexa+ MCP Server)

SmartCommute is a real-time Model Context Protocol (MCP) Server and interactive Alexa+ Smart Display Simulator built for the Build, Ship, Shape: Amazon Developer Hackathon (Alexa+ Track).

🎥 Video Demo

🔗 Watch the Full Demo Video Here: [PASTE YOUR YOUTUBE OR LOOM VIDEO LINK HERE]

Related MCP server: Beckn Mobility MCP Server

💡 The Problem

Every day, commuters waste time switching between 3 to 4 separate apps—Google Maps for traffic, Uber/Ola for cab fares, Rapido for bike taxis, and Metro apps for train timings—just to answer one simple question:

"How do I reach my destination before my deadline without overspending?"

🚀 The Solution

With SmartCommute, a user asks a single natural-language question to Alexa+ or any MCP-enabled AI agent:

"Alexa, I need to reach Z Square Mall from IIT Kanpur in 35 minutes. What is my fastest and cheapest option?"

The AI agent invokes the find_best_commute tool on our live HTTP MCP server, which executes a 4-stage pipeline in real time:

1. Live GPS Geocoding: Converts any real-world origin and destination into exact latitude/longitude coordinates via the OpenStreetMap Nominatim API.

2. Real Road Routing: Fetches the actual road network distance (in km) and live driving duration via the OSRM (Open Source Routing Machine) API.

3. Multimodal Fare & Time Engine: Computes travel times and dynamic INR (₹) fares across four Indian transit modes:

   • 🛵 Bike Taxi (Rapido): Cuts through urban traffic congestion directly.

   • 🚇 Metro Rail: High-speed transit corridor calculation with station walk buffer.

   • 🚕 AC Cab (Uber / Ola): Full road distance with city traffic multiplier.

   • 🚌 City AC Bus: Economical public transit factoring in scheduled stop delays.

4. Deadline Filtering & Voice Synthesis: Filters out transit modes that will make the user late, returning both a structured JSON payload for visual cards and a conversational voice response for Alexa+.

🏛️ System Architecture

Protocol Flow:

1. Client Request $\rightarrow$ POST http://localhost:8000/mcp (method: "tools/call", name: "find_best_commute")

2. MCP Server $\rightarrow$ Queries nominatim.openstreetmap.org & router.project-osrm.org

3. Decision Engine $\rightarrow$ Evaluates Bike Taxi, Metro Rail, AC Cab, and City AC Bus against max_time_minutes

4. Strict MCP Response $\rightarrow$ Returns JSON-RPC 2.0 result.content back to the AI Agent & Smart Display UI

📂 Repository Structure

• server.py — Multi-threaded, zero-dependency Python MCP Server implementing MCP Spec 2025-11-25+ (initialize, tools/list, tools/call).

• index.html — Interactive Alexa+ Smart Display web simulator with live voice synthesis (window.speechSynthesis).

• mcp.json — Configuration file for connecting the HTTP MCP server directly to VS Code Copilot / MCP clients.

• requirements.txt — Environment documentation (runs natively on Python 3.8+ standard library with zero external C++ build dependencies).

• LICENSE — MIT Open Source License.

⚙️ How to Run Locally (Quick Start)

Step 1: Clone the Repository & Start the MCP Server

No pip install is required! server.py uses Python's standard library for instant cross-platform execution:

git clone [https://github.com/AnantM2007/alexa-smart-commute-mcp.git](https://github.com/AnantM2007/alexa-smart-commute-mcp.git)
cd alexa-smart-commute-mcp
python server.py

The server will start at: http://localhost:8000/mcp

Step 2: Test with an AI Agent (VS Code MCP Client)

Move mcp.json into a .vscode/mcp.json folder in your workspace and ask Copilot Chat in Agent Mode:

"Using my MCP server, tell me the cheapest mode of transport from IIT Kanpur to Z Square Mall Kanpur."

Step 3: Launch the Alexa+ Smart Display Simulator

Open index.html in any modern browser, enter any two real-world locations and your deadline in minutes, and click Ask Alexa+.