LPU students develop AI-powered drone detection system to combat cross-border threats

Students from Lovely Professional University (LPU) have developed a cutting-edge, AI-powered drone detection system designed to tackle the rising misuse of drones for cross-border smuggling, surveillance and unauthoried intrusion.

Advertisement

The pioneering project was led by Lakshit Gupta, Vansh Sharma, and Riya Singh, students from the School of Computer Science and Engineering (Machine Learning Domain). The trio worked under the guidance of Dr Mohit Arora, HoD of the Machine Learning Domain, to build a system that detects drones in real time using only a simple camera and artificial intelligence — offering a low-cost alternative to expensive radar and RF-based systems.

“We were driven by the constant reports of drones being used for illegal cross-border activities,” said Lakshit Gupta. “We realised there was a need for an accessible solution that didn’t require high-end, military-grade hardware.”

The system uses live video feeds processed through state-of-the-art AI models such as YOLOv8s, RetinaNet, and One-Class SVM. These models are trained to accurately recognise drones across different environments and lighting conditions. Once a drone is detected, the system can instantly trigger alerts, making it highly effective for real-time monitoring. Further enhancements are in progress to estimate the drone’s position, speed, and altitude.

Though initially designed for defence and border control, the team emphasised that the system has wide-reaching potential. It can be used at airports, public events, industrial facilities, and even in agriculture field — where drone intrusion poses a risk.

“This is a shining example of how LPU students are solving real-world problems using emerging technologies,” said Dr Arora. “The project demonstrates the power of AI to deliver scalable, cost-effective solutions for both security forces and civilians.”

As the threat of unauthorised drone activity continues to grow, this innovative system developed at LPU stands out as a promising tool for enhancing airspace security nationwide.