Closed-loop speed control for a biohybrid neural interface using frequency-modulated lateral-line stimulation

OBJECTIVE: To achieve robust and energy-efficient target-speed regulation in an aquatic biohybrid system by integrating closed-loop feedback with frequency-modulated lateral-line neuromodulation in free-swimming eels. Approach. We present a closed-loop controlled eel-computer biohybrid system that modulates the swimming behavior of Monopterus albus via electrical stimulation of lateral-line nerves. A lightweight microelectronic backpack with stimulation electrodes was noninvasively attached to the eel to enable real-time motion control. Monophasic pulses with optimized parameters (anterior lateral-line nerves, ALLNs: 4.0 V; posterior lateral-line nerves, PLLNs: 2.5 V; 10-100 Hz; 30% duty cycle) elicited directional locomotion. A proportional-derivative (PD) feedback controller dynamically adjusted the stimulation frequency based on real-time velocity measurements to regulate predefined swimming speeds. Main results. PLLNs stimulation induced forward swimming at 9.0-17.8 cm/s, exhibiting a strong negative correlation between speed and frequency (r = -0.96). Conversely, ALLNs stimulation produced backward swimming at 6.9-9.5 cm/s with a moderately strong negative correlation (r = -0.89). Compared with fixed-frequency stimulation, the closed-loop scheme reduced the velocity deviations by 63.7% (forward) and 50.6% (backward). Body-mass-normalized external stimulation power ranged from 0.87 to 1.54 mW/g. Significance. These findings demonstrate bidirectional closed-loop speed regulation in freely swimming eels and provide a compact, noninvasive framework for adaptive, energy-efficient aquatic biohybrid robotic systems based on lateral-line-driven neural interfaces. .