National Winner (Hardware Edition)

Smart India Hackathon 2025

Coast Scan: A decentralized maritime surveillance system developed under 36 hours.

Team

Team Lead

Tech Stack

Raspberry Pi, OpenCV, LoRa

Constraint

36 Hours

Prize

₹1,00,000

Government Problem Statement

The Ministry of Defence sought a low-cost, scalable solution to monitor small fishing vessels that often go undetected by coastal radar, posing security risks.

System Architecture

We designed a network of solar-powered "Smart Buoys." Each buoy operates autonomously:
1. Visual AI: An onboard camera feeds images to a quantized MobileNet model running on a Raspberry Pi.
2. Mesh Network: Buoys communicate with each other and a base station using LoRa (Long Range) radio, which requires no cellular infrastructure.

Team Engineering

As Team Lead, I managed the integration of hardware and software. We split into two sub-teams:
- Hardware: Focused on waterproofing and power management.
- Software: Focused on optimizing the inference model to run at 15 FPS on limited compute.

Why It Won

The judges cited our "pragmatic approach to connectivity." While other teams relied on 5G (which is non-existent 10km offshore), our use of LoRa demonstrated a deep understanding of the deployment environment.

Learnings

Winning SIH taught me that the best engineering solution isn't always the most high-tech one—it's the one that actually works in the field constraints. It was a masterclass in compromise and optimization.