Southeast University team builds planar nanophotonic motion sensor for wide-angle tracking

By AI, Created 5:23 PM UTC, May 18, 2026, /AGP/ – Researchers at Southeast University have developed a bioinspired planar nanophotonic sensor that uses metalens arrays to detect motion across a 135-degree field of view. The system is designed to improve perception and trajectory prediction for tiny, slow-moving targets in cluttered environments, with potential uses in robotics, industrial safety and unmanned systems.

Why it matters: - The sensor tackles a common weakness in machine vision: reliably spotting and tracking small, slow-moving objects in cluttered scenes. - The planar design is thinner, lighter and easier to integrate than traditional curved bionic eye systems. - The approach could support autonomous perception in mobile platforms, industrial early warning and unmanned intelligent systems.

What happened: - A research team led by Prof. Ji Chen at Southeast University proposed and demonstrated a bioinspired motion detector based on planar metalens arrays. - The work appeared in Light: Advanced Manufacturing in a paper with DOI 10.37188/lam.2026.064. - The system achieves an ultra-wide 135-degree field of view. - The detector extracts velocity and direction information from moving targets. - The system also predicts subsequent motion trajectories.

The details: - The motion detector combines a planar metalens array with an optical filter, stray light mask and CMOS image sensor in an integrated module. - The team built a PSF-based convolution imaging model to generate high-fidelity datasets for training and evaluation. - The researchers developed a multi-scale motion perception network. - The researchers also created a lightweight trajectory prediction framework. - The sensor is designed to identify tiny and slow-moving objects in complex backgrounds. - The system is intended to work on mobile platforms such as micro unmanned vehicles. - The research team said the design uses a 1×3 metalens array to enable planar ultra-wide-angle imaging. - The team said the imaging model addresses the lack of dedicated datasets for metalens vision. - The team said the neural network and prediction framework support high-precision perception and multi-target trajectory forecasting under interference backgrounds. - The researchers said the lightweight algorithm architecture and high hardware integration enable millisecond-level trajectory prediction. - The researchers said the system can stably track multiple crossing and overlapping targets. - The project was supported by the National High-Level Personnel of Special Support, Basic Research Program of Jiangsu, Young Elite Scientists Sponsorship Program by CAST, National Natural Science Foundation of China, Jiangsu Key R&D Program Project, Fundamental Research Funds for the Central Universities and the Project of the National Mobile Communications Research Laboratory.

Between the lines: - The work reflects a broader push to replace bulky curved optics with flat nanophotonic hardware that can be more easily embedded in compact devices. - The modeling and neural network components matter as much as the optics, because reliable dataset generation and prediction are key to practical deployment. - The combination of wide-angle imaging and motion forecasting suggests the sensor is aimed at real-world edge cases, not just controlled lab scenes.

What’s next: - The researchers see the platform as a route toward next-generation miniaturized intelligent photoelectric sensing. - The team forecasts applications in robotic vision, vehicle environmental perception, industrial safety warning, low-altitude security monitoring and autonomous decision-making for unmanned systems. - Further work will likely focus on moving the design from demonstration toward deployment on real platforms and in more complex environments.

The bottom line: - Southeast University’s planar metalens sensor pairs wide-angle optics with motion prediction to improve how small targets are detected and tracked in difficult settings.