DRISHTI

Overview

DRISHTI is an off-grid, multi-node intrusion detection system built for dense forest deployment. Distributed field nodes fuse multiple sensing modalities to detect motion, and a machine learning model classifies each event as human, animal or environment, so that real threats can be separated from everyday forest activity.

Detections are relayed over a long-range wireless link to a base station that runs the classifier, raises a siren on high-confidence human detections and pushes live alerts, a map view and a speed graph to a companion Android app.

Components & Hardware

Confidential. The component list and hardware details of DRISHTI are proprietary to Nura Net Space and are not publicly disclosed. For technical or partnership enquiries, please use the Send a Query button.

Design & Engineering Process

DRISHTI is being built through iterative bring-up: every subsystem is tested on its own, then integrated and validated before anything is committed to a final layout.

• Designing a four-layer architecture: field nodes, transmitter station, long-range link and base station with app.
• Developing node firmware that fuses sensor readings and streams each event as structured records.
• Building and validating the communication links through bench and range tests.
• Training machine learning models to classify events as human, animal or environment.
• Running the base station server as always-on services with secure remote access.
• Building the Android app with live map, alerts and notifications.
• Collecting and labelling training data across classes, distances and conditions, followed by full end-to-end testing and forest deployment.

Applications & Use Cases

DRISHTI is designed to give remote, unpowered terrain a low-cost early-warning perimeter that runs without mains power or mobile coverage.

• Forest and wildlife area intrusion detection
• Anti-poaching and illegal logging surveillance
• Remote border and perimeter monitoring
• Off-grid security for farms, orchards and protected sites
• Research platform for sensor fusion and ML-based motion classification