HAWKNET : LONG RANGE AI BASED IOT SYSTEM FOR BORDER SECURITY
by Sharat Singh, Computer Vision Lead
1. Introduction
This project was aimed for a hackathon organized jointly by the The Ministry of Electronics and Information Technology, Government of India (MeitY) and Sashtra Seema Bal (SSB), which is one of India's Central Armed Police Forces, and is responsible for guarding the Indo-Nepal and Indo-Bhutan borders
2. Problem Statement
Detection of unauthorized/suspected vehicles and persons moving across the border using AI based IoT system
The Problem Statement was complex since the product was intended for SSB (Sashastra Seema Bal), who are peacekeeping forces deployed along the Nepal and Bhutan borders, which brings in a whole new dimension of focus factors like terrain, security, encryption, robustness, performance and maintenance.
3. Solution
A fleet of standalone IoT devices to detect human/vehicle movement in unauthorized areas and store them in database
The module is made up using the following components -> NVIDIA Jetson Nano, Arducam 8MP PTZ, LoRa 868MHZ SX1276
The following Technical Stack is used: Python, Computer Vision, Machine Learning, IoT Communication Protocol LoRa WAN, AWS Cloud Services
4. Data Generation
The data generation was done using a combination of Computer Vision and Machine Learning. This data would then be sent to the central server through nodal devices using LoRaWAN
Machine Leaning model and image processing done onboard, and sent to Gateway using LoRaWAN, therefore will work in areas without 3G/4G/5G network
Battery backup paired with Solar Panels will run the Standalone device estimated ~1 year without maintenance
Want to talk or discuss something?
More projects
AGROBOT : AGRICULTURE BASED ROBOT WITH SMART VISION
This is a robot simulation done in ROS using the Gazebo Simulator for use cases in the agricultural industry, bridging the gap between the on ground farmers and the agricultural scientists, researchers and student with the aim of smartly moving across a specified map, and detecting fruits/vegetables that are ripe/consumer-ready using computer vision and AI, and retrieving it using a mechanical grabbing arm
DEPAM: DATA EXTRACTION, PROCESSING, AUTOMATION and MANIPULATION
This is a part of my Dissertation Project for my Bachelors Degree titled 'TCAD SIMULATION AND INVESTIGATION OF FERROELECTRIC FIELD EFFECT TRANSISTORS (FeFETs) FOR MEMORY APPLICATIONS WITH KEYSIGHT ADS MODEL USING VERILOG-A', and deals with the automation of the data extraction and processing from the TCAD simulation files, and the manipulation of the data to be used in the Verilog-A model