Project Title: Ishtar Guardian Eye: A Multi-Modal Sensor Fusion Platform for Autonomous Drone Detection and Tracking
Description:
Problem Statement:
The rapid proliferation of low-cost drones poses significant security risks for sensitive areas, including airports, industrial facilities, and critical infrastructure. Conventional surveillance systems, relying on single-sensor modalities like cameras or radar, often fail to detect small, agile UAVs, especially in low-light, occluded, or noisy environments. High false-positive rates, limited coverage, and susceptibility to mechanical or environmental interference highlight the urgent need for a robust, real-time, and accurate detection system.
Proposed Solution / Innovation:
The Ishtar Guardian Eye introduces a multi-modal sensor fusion platform, integrating computer vision (YOLOv8), 24 GHz mmWave radar, RF scanning, and directional acoustic sensing. By combining complementary sensing modalities, the system achieves early target detection, continuous tracking, and significantly reduced false positives. A decentralized architecture separates AI computation on an NVIDIA RTX 4060 GPU from real-time mechanical control via an ESP32 microcontroller and high-torque Pan-Tilt servos. This hybrid design enables cost-effective, high-performance surveillance comparable to military-grade systems, while remaining scalable and deployable on consumer-grade hardware.
Method / Implementation:
The system operates in a layered process: radar and acoustic sensors detect potential targets, which are validated by AI-based visual detection. Detected drones are continuously tracked using a CSRT algorithm, while positional errors are corrected via a closed-loop control system that drives the Pan-Tilt mechanism. Software is developed in Python for AI and sensor fusion tasks, and C++ for embedded control on the ESP32. Real-time communication between components is maintained through serial protocols, ensuring stable and responsive tracking.
Key Components (HW/SW/AI):
Hardware: NVIDIA RTX 4060 GPU, ESP32 microcontroller, HLK-LD2410 mmWave radar, MG996R Pan-Tilt servos, Directional Shotgun Microphone.
Software: Python 3.10, C++ firmware, OpenCV, PySerial, CUDA-enabled deep learning libraries.
AI Models: YOLOv8 for drone detection, CSRT for persistent tracking, FFT-based acoustic signal processing.
Results / Impact:
The system achieves real-time drone detection with <30 ms latency, maintains a tracking error margin of ±5 cm, and significantly reduces false-positive rates compared to single-sensor systems. Its multi-modal approach ensures reliable operation in low-light, occluded, and noisy environments. The platform provides scalable, cost-effective protection for airports, industrial sites, and urban infrastructure, improving situational awareness and response time.
Demo Summary (60s Video):
The video introduces the Ishtar Guardian Eye, an autonomous multi-sensor fusion system designed for advanced drone defense
. The system addresses evolving invisible aerial threats by providing a hybrid surveillance solution that merges computer vision, high-precision radar, and RF scanning into a single unified perimeter.
Key technical highlights of the demonstration include:
Processing Power: The system is fueled by the RTX 4060 and utilizes YOLO V8 algorithms to analyze thousands of data points in milliseconds.
Tracking Capabilities: It offers instant detection and persistent trajectory tracking, maintaining a complete target lock even during high-speed or complex maneuvers.
Situational Awareness: The suite provides situational awareness and is designed to operate flawlessly in the most challenging environments where traditional surveillance might fail.
Uploaded At: 2026-03-23 04:11:40