Azernews.Az

By Alimat Aliyeva

An international team of researchers has developed a neuromorphic machine vision system capable of reacting faster than the human eye, significantly enhancing the safety of autonomous driving, AzerNEWS reports.

Traditional autonomous vehicle systems often struggle with response delays. For instance, a self-driving car traveling at 80 km/h may take about 0.5 seconds to detect and recognize an obstacle, whereas a human driver typically reacts in around 0.15 seconds. During that half-second delay, the vehicle can travel an additional 13 meters — a distance that can pose serious safety risks in critical situations.

The newly developed neuromorphic system accelerates visual data processing to speeds up to four times faster than human reaction time, while maintaining or even improving accuracy. Importantly, the technology does not replace conventional cameras. Instead, it integrates with existing hardware through specialized plug-in modules, enhancing the performance of standard computer vision algorithms.

The system is inspired by the way the human eye works. Rather than processing every detail in a scene, it focuses only on moving objects and ignores static background elements, significantly reducing the amount of data that must be analyzed. Its hardware is based on two-dimensional synaptic transistors — highly sensitive motion-detection chips capable of capturing image changes within just 100 microseconds. These chips can retain motion information for more than 10,000 seconds and endure over 8,000 operating cycles without performance degradation.

In laboratory tests, the system demonstrated up to a tenfold acceleration in visual data analysis compared to traditional methods. Beyond autonomous vehicles, the technology shows strong potential for applications in robotics, drones, industrial automation, and smart surveillance systems.

Researchers emphasize that this breakthrough could mark an important step toward safer and more efficient autonomous technologies, reducing accident risks and improving real-time decision-making in complex environments.