Gazzola's soft bendable CyberOctopus called a machine learning marvel

Octopuses have a distributed nervous system, with a brain in their mantle handling complex tasks such as learning and decision-making, and neural tissue in their limbs managing simpler motor commands. 

Inspired by this hierarchical brain structure, a team led by MechSE professor Mattia Gazzola built the “CyberOctopus,” a simulated multi-limbed soft robot that harnesses a hierarchical machine learning strategy to forage for virtual treats. 

In their study published in Advanced Intelligent Systems, they presented the CyberOctopus, a simulated multi-limbed soft robot that harnesses a hierarchical machine learning strategy to forage for virtual treats.

Read the feature on Advanced Science News >>