Multimodal spiking neural network for space robotic manipulation

doi:10.1016/j.actaastro.2026.04.049

Multimodal spiking neural network for space robotic manipulation

2026-04-22

0

PRE

AI

DOI：10.1016/j.actaastro.2026.04.049

Original

Original request for help

Save

Abstract

En 中文

• Unified architecture controls space robots across diverse mounting setups. • Multimodal framework fuses geometric and tactile data for robust grasping. • Curriculum learning achieves high manipulation success in microgravity. • Spiking neural networks use far less energy than artificial neural networks.

Keywords:

space robots

multimodal framework

spiking neural networks

robotic manipulation

curriculum learning

Multimodal spiking neural network for space robotic manipulation

Abstract

Journal

Organization