A Brain-inspired Embodied Intelligence for Fluid and Fast Reflexive Robotics Control

TL;DR

This paper introduces NeuroVLA, a bio-inspired neuromorphic framework for robotic control that achieves rapid, stable, and reflexive actions with minimal energy, mimicking biological nervous system functions.

Contribution

The paper presents the first neuromorphic vision-language-action system deployed on physical robots, demonstrating biological motor traits and real-time reflexes without extra data.

Findings

Achieves stable robotic arm control and safety reflexes in under 20 ms

Consumes only 0.4W energy on neuromorphic hardware

Exhibits biological-like temporal memory and reflex responses

Abstract

Recent advances in embodied intelligence have leveraged massive scaling of data and model parameters to master natural-language command following and multi-task control. In contrast, biological systems demonstrate an innate ability to acquire skills rapidly from sparse experience. Crucially, current robotic policies struggle to replicate the dynamic stability, reflexive responsiveness, and temporal memory inherent in biological motion. Here we present Neuromorphic Vision-Language-Action (NeuroVLA), a framework that mimics the structural organization of the bio-nervous system between the cortex, cerebellum, and spinal cord. We adopt a system-level bio-inspired design: a high-level model plans goals, an adaptive cerebellum module stabilizes motion using high-frequency sensors feedback, and a bio-inspired spinal layer executes lightning-fast actions generation. NeuroVLA represents the first deployment of a neuromorphic VLA on physical robotics, achieving state-of-the-art performance. We observe the emergence of biological motor characteristics without additional data or special guidance: it stops the shaking in robotic arms, saves significant energy(only 0.4w on Neuromorphic Processor), shows temporal memory ability and triggers safety reflexes in less than 20 milliseconds.