Recent advancements in optoelectronic and photonic technologies have opened new possibilities for high-speed control applications. These systems operate at significantly higher frequencies than traditional electronics, enabling ultrafast signal processing with advantages such as low latency, high bandwidth, and resistance to electromagnetic interference. Additionally, the integration of neuro-inspired learning mechanisms offers a novel approach to adaptive control. Unlike conventional methods that rely on fixed parameters, these learning-based techniques allow real-time adjustments based on environmental changes. This shift toward adaptive, high-speed control has the potential to enhance performance across various dynamic and rapidly changing systems.

Researchers Silvia Ortín, Moritz Pflüger, and Apostolos Argyris from the Institute for Cross-Disciplinary Physics and Complex Systems (IFISC) in Palma de Mallorca, Spain, are dedicated to interdisciplinary problem-solving by combining bio-inspired computing with high-speed optoelectronics. By integrating optical components with Hebbian learning, their work bridges photonics and machine learning, unlocking new possibilities for ultra-fast, bio-inspired computational systems. A significant innovation in their research is the feedback mechanism in which the physical properties of the system autonomously adapt based on an input correlation (ICO) learning rule. This represents a major step toward creating self-sustaining optoelectronic platforms. In their prototype, they presented a fiber-based dendritic structure with adaptive plasticity for autonomous learning and control, leveraging such Hebbian principles. Central to their computational framework is a closed-loop controller embedded within the experimental dendritic unit, capable of real-time operation at 1 GHz signaling and sampling rates. This system was successfully applied to a hypothetical temperature stabilization task, demonstrating its potential for ultra-fast, adaptive control. The work entitled “Temperature stabilization with Hebbian learning using an autonomous optoelectronic dendritic unit” was published on Frontiers of Optoelectronics (published on April 3, 2025).
DOI: 10.1007/s12200-025-00151-9