Criticality-Driven Evolution of Adaptable Morphologies of Voxel-Based Soft-Robots

Area of Science:

• Robotics
• Complex Systems
• Morphological Computation

Background:

• Voxel-based soft robots shift complexity from control to morphology.
• Adaptability in soft robots, the ability to handle diverse tasks, is crucial but not fully understood.
• Current design methods for adaptable morphologies are limited.

Purpose of the Study:

• To propose a task-agnostic method for designing adaptable soft robotic morphologies.
• To investigate the concept of criticality as a guiding principle for morphology design.
• To validate if criticality correlates with enhanced robot adaptability.

Main Methods:

• Developed a criticality measure for voxel-based soft robots using avalanche analysis.
• Employed an evolutionary approach to guide robot morphologies towards critical dynamics.
• Assessed robot performance on three distinct tasks requiring different skills.

Main Results:

• Morphologies evolved towards criticality showed improved adaptability.
• The criticality measure effectively predicted a morphology's ability to cope with varied tasks.
• Critically tuned robots performed comparably to or better than handcrafted designs.

Conclusions:

• Criticality is a key factor for achieving adaptable soft robotic morphologies.
• This approach offers a promising, task-agnostic method for designing versatile voxel-based robots.
• Further research can leverage criticality for advanced soft robot design.