Sensorimotor Model of Obstacle Avoidance in Echolocating Bats.

Area of Science:

• Bioacoustics
• Computational Neuroscience
• Animal Behavior

Background:

• Bat echolocation involves complex subtasks like navigation and prey detection.
• Identifying minimal acoustic cues for each echolocation task is crucial for understanding bat capabilities.
• Minimal cues for obstacle avoidance in bats remain largely unknown.

Purpose of the Study:

• To propose and computationally validate that Interaural Intensity Difference (IID) and first millisecond echo travel time are sufficient cues for bat obstacle avoidance.
• To present a control algorithm for obstacle avoidance based on these proposed cues.
• To investigate the applicability of these cues for both constant frequency (CF) and frequency modulated (FM) bats.

Main Methods:

• Spatial simulations in 2D and 3D environments were used to model bat echolocation and obstacle avoidance.
• A control algorithm incorporating IID, echo travel time, and alternating ear positions was developed.
• Simulations were performed using models inspired by Rhinolophus rouxii (CF bat) and Phyllostomus discolor (FM bat).

Main Results:

• The proposed algorithm successfully steered simulated bats clear of obstacles in complex environments.
• Obstacle avoidance was achieved without requiring the bat to reconstruct a 3D map of its surroundings.
• The findings suggest that both CF and FM bats could potentially utilize these minimal acoustic cues.

Conclusions:

• Interaural Intensity Difference (IID) and early echo travel time are sufficient acoustic cues for bat obstacle avoidance.
• Simple phonotaxis based on these cues allows for real-time navigation around obstacles.
• This computationally explicit model provides a validated explanation for a fundamental bat behavior, applicable across different bat types.