Neuromorphic Infrared Fibre-Optic Event-Based Sensing with Fast and Efficient Photonic-Electronic Spiking Neurons
This paper presents a novel neuromorphic infrared fibre-optic sensing system that utilizes photo-detecting resonant tunnelling diodes as light-triggered spiking neurons to enable fast, energy-efficient, and event-driven remote detection of environmental phenomena like temperature, motion, and audio, thereby overcoming the data redundancy and processing limitations of traditional photonic sensing.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the modern world, our sensors are often overwhelmed by the sheer volume of data they collect. Whether monitoring a bridge for cracks, tracking weather patterns, or listening for intruders, traditional devices capture a continuous stream of information, recording every moment whether anything important is happening or not. This constant recording consumes vast amounts of energy and requires massive storage systems, creating a bottleneck that slows down decision-making. Nature, however, has solved this problem long ago. Biological senses, like our own eyes and ears, do not record a steady stream of data; instead, they operate on an event-driven basis. They remain quiet until something changes, and then they fire a quick signal to the brain only when a new event occurs. This approach is incredibly efficient, using minimal energy to convey exactly what matters. Scientists have long sought to build machines that mimic this biological efficiency, creating "neuromorphic" systems that process information like a brain rather than a standard computer. The challenge has been to combine this smart, event-based logic with the speed and reach of light-based communication, which is the backbone of our global internet infrastructure.
A team of researchers at the University of Strathclyde and their colleagues in Portugal and the UK have now demonstrated a new way to do exactly this. They have built a sensing system that uses standard fiber-optic cables, the same kind that carry internet traffic across oceans, to detect environmental changes. Instead of sending a continuous video feed of data back to a central computer, their system waits for a specific event to happen. When a change occurs—such as a shift in temperature, a vibration, or a sound—it triggers a tiny, ultra-fast electrical spike. This spike acts as a signal that an event has happened, carrying information about the event's intensity without the need to transmit a massive file of raw data. The core of this innovation is a special component called a photo-detecting resonant tunneling diode. Think of this device as a highly sensitive light switch that, when triggered by a specific amount of light, snaps into a state where it fires rapid electrical pulses. By connecting this device to a sensor embedded in a fiber-optic cable, the researchers created a system that can "listen" to the environment and respond with lightning speed, all while using very little power.
The researchers tested their system by embedding a special sensor, known as a Fiber Bragg Grating, directly into the fiber-optic cable. This sensor is designed to reflect light back toward the detector when it is disturbed. In their first experiment, they monitored temperature changes. They warmed the sensor slightly, which caused it to reflect more light back to the detector. As the temperature rose, the detector did not just send a steady signal; it began to fire electrical spikes. The hotter it got, the faster these spikes fired. This allowed the system to distinguish between a resting state and a warming environment instantly. The researchers then tested the system's ability to detect physical movement. By gently tapping the fiber-optic cable, they created small strains that shifted the sensor's reflection. The system responded immediately, firing a burst of spikes for the duration of the tap. It could even distinguish between a light tap and a harder one, as the strength of the tap changed the frequency of the spikes.
Moving beyond simple taps and temperature, the team demonstrated that the system could detect sound and even fast-moving air currents. They attached the fiber-optic cable to a speaker playing audio tones. As the speaker vibrated, the cable strained, and the detector fired spikes in rhythm with the sound waves. When they played complex sounds, such as the crackling of a fire or the roar of an airplane, the system produced unique patterns of spikes for each sound, effectively encoding the audio into a stream of electrical events. Finally, they tested the system against fast-moving turbulence by blowing compressed air at the cable. The system detected these rapid, chaotic movements, firing spikes at rates reaching up to 11 million times per second. This proved that the technology could handle events changing thousands of times per second, a speed far beyond what many traditional sensors can manage efficiently.
The significance of this work lies in its ability to bring intelligence to the edge of the network. Instead of sending terabytes of raw data to a central server for analysis, the sensor itself processes the information and sends only the essential events. This approach drastically reduces the energy required for monitoring and the storage space needed for data. The researchers showed that their system works across a wide range of speeds, from slow temperature shifts that happen over minutes to rapid air turbulence that changes in milliseconds. Because the system uses standard fiber-optic technology and operates at the wavelengths used for global telecommunications, it can be integrated into existing infrastructure without needing a complete overhaul. The study confirms that it is possible to create remote sensing networks that are not only fast and efficient but also capable of making immediate decisions based on what they "see" and "hear." This opens the door for smarter monitoring of critical infrastructure, from bridges and buildings to security perimeters, where quick, energy-efficient detection of events could prevent disasters and improve safety.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.