Visually guided swarm motion coordination via insect-inspired small target motion reactions.

Area of Science:

• Robotics
• Neuroscience
• Control Systems

Background:

• Insect sensory systems offer bio-inspired models for robotic agent coordination.
• Existing bio-inspired approaches lack systematic understanding for multi-agent robotic swarm performance.
• Barriers exist in implementing bio-inspired sensing and feedback for robotic swarms.

Purpose of the Study:

• Introduce the small-target motion reactive (STMR) swarming approach for robotic applications.
• Develop a concise engineering model of insect small target motion detector (STMD) neurons.
• Analyze the stability and group performance of the STMR approach.

Main Methods:

• Designed an engineering model of insect STMD neurons to detect peak optic flow.
• Developed an output feedback switched control system based on STMD neuron output.
• Conducted theoretical stability analysis for bi-agent and group contexts.
• Validated the STMR approach through simulations and ground vehicle implementations.

Main Results:

• The STMR approach achieves collective group motion despite minimal agent connectivity (single neighbor).
• Theoretical analysis confirms bi-agent stability and state boundedness in group settings.
• Simulations and robotic implementations validate the STMR approach's effectiveness.
• Group-level analysis shows continuously varying polarization and decreasing heading variance.

Conclusions:

• The STMR approach provides a viable bio-inspired method for robotic swarm coordination.
• Minimal connectivity does not hinder effective collective motion in STMR swarms.
• STMR enhances swarm behavior through polarization and heading control, applicable to robotic systems.