Can robots without Hebbian plasticity make good models of adaptive behaviour?

Area of Science:

• Biorobotics
• Neuroscience
• Animal locomotion

Background:

• Current biorobotic models may not adequately capture the complexities of animal movement control.
• Animals primarily solve movement shaping problems using sensory guidance and adaptation.

Purpose of the Study:

• To propose a refined framework for developing biorobotic models that accurately represent animal movement.
• To emphasize the necessity of biological realism in biorobotic system design.

Main Methods:

• Defining constraints for biorobotic model variables and mechanisms.
• Specifying requirements for input vectors to model sensor fields.
• Detailing the need for internal state vectors to model neurophysiological processes.
• Ensuring output vectors accurately represent coordinated muscle signals.

Main Results:

• Biorobotic models require real-world counterparts for all variables and mechanisms.
• Input data must accurately reflect known sensory fields.
• Internal model states and transformations should mirror neurophysiological processes.
• Output signals must precisely model coordinated muscle activity.

Conclusions:

• A stricter class of biorobotic models is needed for accurately simulating animal movement.
• Biological fidelity in sensory, internal, and motor components is crucial for effective biorobotic systems.