Smart Air Defense System for Real-time Threat Detection
Defence Software and a Swarm of Sensors Powered by AI for Real-time Sound Detection and Response for Airborne Threats
Goal
Defending Europe's territory from external and internal threats utilizing preventive means and AI technologies
Project duration
June 2022 - ongoing
SERVICES WE PROVIDED
Challenges in Defence System
The unexpected appearance of air threads
It can be challenging to detect aerial targets that fly at extremely low altitudes and navigate through river valleys, lowlands, and ravines while avoiding populated areas without being noticed.
The margin of error and delays of traditional defense software
Traditional air defense systems often face challenges in accurately tracking and providing real-time updates on trajectories and air threat statuses.
High flight speed of the air threads
Due to the high speed of flight of some air targets, their type can be difficult to determine by observation without expensive equipment.
Solution
By deploying advanced threat detection and response solutions that consist of audio surveillance network in vulnerable areas, we can constantly monitor the environment for any unusual sound patterns.
This allows us to identify potential threats with exceptional accuracy and speed. We’ve tested the concept and want to start designing a production-ready device.
CORE FEATURES WE IMPLEMENTED
Real-time Threat Sound Identification
Integration with Defense Centers
Data Analysis and Categorization of Sound
Protection Against Moisture and Dust
Leveling the Human Factor
Determining the Position of the Target
Determining the Flight Trajectory, Plotting It on the Map
Sound and Noise Filtering Using NNL
Low Manufacturing Cost
TEAM COMPOSITION
1 Project Manager
1 Tech Architect
1 Hardware Designer
2 AI Solution Engineers
TECHNOLOGY STACK
The details of the technology used are confidential and protected by a non-disclosure agreement (NDA).RESULTS WE ACHIEVED
- A sensor prototype has been developed
- A dataset for neural network training is collected
- The neural network is trained to detect cruise missiles and fighter jets
After testing the concept, the team intends to proceed with designing and perfecting a device that is ready for production.
