bhagyareddie55/METAHACKATHON

🚖 Smart Ride Suggestion API

📌 Overview

The Smart Ride Suggestion API is a backend system designed to simulate intelligent ride selection based on availability, ETA (Estimated Time of Arrival), and realistic distance. Instead of automatically choosing one service, the system provides neutral recommendations and lets users decide — similar to real-world platforms like Uber, Ola, and Rapido.

Project Structure BHAGYA/ │── app.py # Main FastAPI application │── ride_env.py # Core ride simulation logic │── env.py # Environment setup (provided template) │── baseline.py # Baseline logic (provided) │── inference.py # Inference handling │── models.py # Data models │── grader.py # Evaluation logic │── tasks.py # Task definitions │── Dockerfile # Container setup │── requirements.txt # Dependencies │── openenv.yaml # OpenEnv configuration

Features Smart ride suggestions based on availability Supports real Bangalore locations (Hebbal, Whitefield, BTM, Yelahanka, etc.) Realistic distance calculation (Haversine + road factor) Dynamic ETA (varies each request) No bias toward any provider Direct booking/deep links for each ride service Supports reset, step, and state APIs Reinforcement-learning inspired environment design

Tech Stack Backend: FastAPI Language: Python Geolocation: Geopy (Nominatim) Server: Uvicorn Containerization: Docker

API Endpoints

1. 1./ GET Returns API status.
2. 2./reset POST Resets the environment.
3. 3./step POST Generates ride suggestions. Request Example: { "pickup": "hebbal", "drop": "whitefield", "radius_km": 5 }

Response Includes: Distance (km) Ride availability ETA for each provider Booking links Recommendation

1. 1./state GET Returns the last computed state.

How It Works

1. 1.User inputs pickup and drop locations

1. 1.Input is cleaned (e.g., “btm” → “BTM Layout Bangalore”)

1. 1.Locations converted to coordinates using Geopy

1. 1.Distance calculated using Haversine formula

1. 1.Adjusted using a factor (1.2–1.4) for realistic road distance

1. 1.Ride availability simulated dynamically

1. 1.ETA generated based on availability + randomness

1. 1.System suggests best options without bias

1. 1.Users can continue booking via provided links

Distance Logic Base: Haversine formula Adjustment: Multiplied by 1.2–1.4 Ensures realistic city travel distances

ETA Logic Higher availability → Lower ETA Lower availability → Higher ETA Random variation simulates real-world conditions

Ride Providers The system supports: 1.Uber 2.Ola 3.Rapido 4.Local services (via Justdial)

Each option includes a direct booking link.

Notes Links appear as plain text in Swagger UI (not clickable) They can be copied and opened in a browser No paid APIs are used (fully self-contained)

Running the Project

1. 1.Install dependencies pip install -r requirements.txt
2. 2.Run the server uvicorn app:app --reload
3. 3.Open in browser http://127.0.0.1:8000/docs

Conclusion This project demonstrates a practical backend system for intelligent ride suggestion using simulation, geolocation, and decision logic. It emphasizes user choice, neutrality, and real-world behavior.