1 of 18

DRISHTI-UAV: AI-Powered Autonomous Surveillance & Rescue Drone

Enhancing disaster response and defence surveillance using intelligent UAV systems.

Objective: Enhance mission efficiency and reduce human risk

Team name: SKYFORGE

Team Members : Harshini R, N Anuja ,Abinaya S

2 of 18

Problem Statement

  • Defence and disaster zones are dangerous and difficult to access

Manual search operations are slow and risky

Poor visibility (fog, rain, night) reduces human detection

Large areas like borders and disaster zones require faster monitoring

      • Existing drones require continuous human monitoring

3 of 18

Motivation & Need

Many disaster victims remain undetected for long periods, reducing survival chances

Rescue teams and soldiers face high risk while searching in dangerous environments

Poor visibility and complex terrains make human detection very difficult

• Traditional surveillance systems rely heavily on manual monitoring, causing delays

AI-powered drones can improve detection speed and support faster rescue operations

Large disaster areas and border regions require faster and more efficient monitoring systems

4 of 18

Proposed Solution DRISHTI-UAV

  • AI-powered multi-sensor drone designed for defence surveillance and disaster rescue missions

• Uses RGB camera and thermal camera to detect humans even in low-visibility conditions

Sensor fusion technology improves detection accuracy by combining multiple sensor data

On-board AI processing enables real-time object detection without cloud dependency

Autonomous search path planning helps the drone cover large areas efficiently

• Sends real-time alerts and location data to the control station for faster response

5 of 18

System Architecture

  1. Perception Hardware (Sensors)RGB Camera – Visual object detection�• Thermal Camera – Detects human body heat in low visibility�• LiDAR / Ultrasonic Sensor – Obstacle detection and avoidance�• GPS Module – Provides location tracking and mapping

2. Intelligence & ProcessingAI object detection using YOLOv8

� • Sensor fusion combines data from multiple sensors for higher accuracy

� • Navigation support using Simultaneous Localization and Mapping (SLAM)

6 of 18

Working Methodology

  • Drone deployed in target region
  • Sensors continuously capture environment data
  • AI analyzes images in real time
  • Detected objects and distances are displayed
  • Alerts sent to defence/rescue teams

WORKFLOW

Mission Start

UAV Deployment

Multi-Sensor Environment Scan

AI Object Detection

Sensor Fusion Verification

Location Mapping (GPS + Depth)

Live Alert to Control Station

Rescue / Defence Response

7 of 18

AI & Technologies Used

The drone is deployed in the target area for surveillance or rescue missions

• Sensors continuously scan the surroundings and collect environmental data

• The onboard AI analyzes the images using YOLOv8 for object detection

• Sensor fusion verifies detection using camera, thermal, and distance data

• The system records GPS location of the detected target

• Real-time alert is sent to the control station for immediate response

8 of 18

Implementation Strategy

  • Hardware Integration: Assemble drone frame, cameras, GPS, and sensors.

  • AI Development: Train and deploy YOLOv8 for human detection.

  • System Testing: Test simulation and real-time detection performance.

  • Field Deployment: Use the drone for surveillance and rescue operations with live alerts.

  • Design & Planning: Define drone architecture, sensors, and AI modules.

9 of 18

PROTOTYPE MODEL & VIDEO

This simulation demonstrates the working of our DRISHTI UAV detection system. An ultrasonic sensor is used to detect objects or humans in the search area. When the sensor detects a person within a certain distance, the system activates an alert using LEDs and sends a signal indicating a possible victim location. This represents how the drone would identify victims during a rescue mission.”

10 of 18

Unique Innovation DRISHTI-UAV

  • AI-Powered Human Detection: Uses deep learning (YOLO) to detect humans and vehicles in real time.

  • Multi-Sensor Fusion: Combines RGB camera,

thermal imaging, and LiDAR/ultrasonic sensors for higher

detection accuracy.

  • Onboard Edge AI Processing: AI runs directly on the drone for instant analysis without cloud delay.

  • Autonomous Search Path Planning: Drone automatically

scans large areas using optimized search patterns.

  • Dual-Use Capability: Designed for both defence Surveillance and Disaster Rescue Operations.

  • Low-Cost Intelligent UAV: Provides advanced AI surveillance at a much lower cost than traditional systems

.

11 of 18

Competitive Comparison

Feature

Traditional Surveillance Drones

DRISHTI-UAV (Proposed System)

Human Detection

Mostly manual monitoring

AI-based detection using YOLOv8

Low Visibility Detection

Limited or not available

Thermal camera detects body heat

Sensor Integration

Single camera system

Multi-sensor fusion (RGB + Thermal + Distance)

Autonomous Operation

Mostly manual control

Autonomous search path planning

Real-Time Alerts

Limited monitoring

Instant alerts to control station

Detection Accuracy

Depends on operator

AI improves detection accuracy

Application Areas

General surveillance

Defence + Disaster Rescue missions

Cost Efficiency

High-cost advanced systems

Low-cost intelligent UAV solution

12 of 18

Advantages

Reduces risk to soldiers and rescue teams by using unmanned aerial surveillance

• Detects humans and vehicles even in low-light, fog, or night conditions using thermal sensing

• Covers large and difficult terrains quickly compared to ground patrol systems

• AI-based object detection enables faster and more accurate monitoring

• Multi-sensor fusion improves reliability by combining camera, thermal, and distance sensors

• Low-cost intelligent UAV system compared to traditional surveillance technologies

13 of 18

Cost Analysis

Project Cost (Estimated)

Drone frame and motors – ₹15,000 – ₹40,000

Flight controller and GPS module – ₹8,000 – ₹20,000

RGB camera and thermal camera – ₹10,000 – ₹80,000

Sensors (Ultrasonic / LiDAR / Weather sensor) – ₹5,000 – ₹20,000

AI processing unit (Raspberry Pi / Jetson Nano) – ₹5,000 – ₹20,000

Total Prototype Cost

₹40,000 – ₹1,80,000 (depending on sensors used)

14 of 18

Target Users

  • Indian Armed Forces

  • Border Security Force

  • National Disaster Response Force (NDRF)

  • Disaster management authorities

  • Search and rescue team

Market Opportunity

  • Defence surveillance market

  • Smart drone industry growth

  • Disaster management market

15 of 18

Business Model

  • Target Customers: Defence forces, Border Security Force, Disaster Response Agencies (NDRF)
  • Value Proposition: AI-powered drone for faster surveillance and rescue operations
  • Revenue Source: Drone system sales + AI software integration + maintenance services
  • Key Benefit: Reduces human risk and improves monitoring efficiency in critical missions

16 of 18

Applications & Use Cases

Border infiltration detection

Post-earthquake survivor detection

Border surveillance and intrusion detection Search & rescue during landslides and floods

Battlefield reconnaissance Disaster damage assessment

Support in GPS-challenged environments

17 of 18

Impact �

  • Faster detection of survivors in disaster zones
  • Reduces risk for soldiers and rescue teams
  • Improves border security monitoring
  • Provides real-time situational awareness
  • Helps save lives during emergencies

Challenges & Solutions

Challenge

Solution

Battery limitation

High-capacity Li-Po battery

Weather conditions

Weather-resistant sensors

False detection

Multi-sensor verification

18 of 18

Conclusion And Future Scope

  • DRISHTI-UAV is an AI-powered drone designed for disaster rescue and defence surveillance.

  • It uses thermal imaging, RGB cameras, and AI detection to identify humans in difficult environments.

  • The system helps reduce risk to rescue teams and improves monitoring efficiency.

  • Future development includes drone swarm systems and long-range communication.

  • It can also support medical supply delivery and advanced AI rescue analysis