Leveling the head relative to ground slope improves visual odometry in simulated bees

Area of Science:

• Behavioral Ecology
• Robotics
• Biomimetics

Background:

• Honeybees (Apis mellifera) use the waggle dance to convey food source information.
• Foragers estimate distance using visual cues like optic flow, even on varied terrain.
• Previous models of the honeybee visual odometer lack robustness on uneven surfaces.

Purpose of the Study:

• To introduce a novel model for the honeybee visual odometer, named Surf.
• To investigate the robustness of optic flow-based visual odometry on uneven terrain.
• To enhance navigation capabilities for aerial robots.

Main Methods:

• Developed the Surf model, integrating vertical oscillations and compound eye reorientation relative to terrain.
• Simulated honeybee visual odometry in open fields with uneven terrain and wind variations.
• Evaluated Surf model performance against previous models under identical conditions.

Main Results:

• The Surf model demonstrated superior accuracy and maintained precision on uneven terrain compared to prior models.
• The model's compound eye reorientation enhances optic flow detection across irregularities.
• Surf model performance aligns with observed honeybee navigation behaviors on slopes.

Conclusions:

• The Surf model provides a more robust estimation of distance for honeybees navigating complex terrain.
• This biologically plausible model offers a reliable framework for aerial robot navigation in challenging environments.
• The Surf model advances our understanding of insect visual odometry and its applications.