A biomimetic vision-based hovercraft accounts for bees' complex behaviour in various corridors.

Area of Science:

• Robotics
• Biomimetics
• Insect Navigation

Background:

• Honeybees exhibit sophisticated visual guidance for navigation.
• Previous studies explored optic flow-based control in simulations.

Purpose of the Study:

• To systematically compare biomimetic robot and honeybee visual guidance.
• To implement and test a novel optic flow-based autopilot on a physical robot.

Main Methods:

• Developed a miniature hovercraft with a 'lateral optic flow regulation autopilot'.
• Utilized two local motion sensors (LMS) for visual input.
• Tested navigation in various corridor configurations and simulated disturbances.

Main Results:

• The robot navigated straight, tapered, and bent corridors safely.
• It reacted appropriately to environmental perturbations like wall texture absence and sloping terrain.
• A minimalistic visual system with lateral optic flow proved sufficient for control.

Conclusions:

• The biomimetic robot's performance validates the optic flow regulation autopilot.
• This system effectively mimics honeybee navigation strategies.
• Simple visual cues can enable robust robotic navigation in complex environments.