Olfactory searches with limited space perception.

Area of Science:

• Computational Biology and Robotics
• Animal Behavior and Navigation

Background:

• Insects and small animals navigate turbulent environments using sparse chemical cues (odor) for locating mates or food.
• The extent of internal space perception and cognitive map usage in these animals remains debated, with limited spatial awareness often suggested.

Purpose of the Study:

• To propose and validate a computational scheme for effective searching in turbulent streams, even with limited or no internal space perception.
• To model animal search behaviors and explore applications in robotic systems for navigation and localization.

Main Methods:

• Developed an algorithm based on standardized projection of source position probability to bypass the need for detailed space perception.
• Incorporated free energy evaluation for path direction and an internal 'temperature' parameter to control exploration-exploitation balance.
• Validated the scheme through numerical simulations of odor plume propagation and experimental robotic searches in turbulent streams.

Main Results:

• The proposed scheme demonstrated efficient searching capabilities in turbulent streams without relying on detailed internal spatial maps.
• Numerical and experimental results confirmed the algorithm's effectiveness in source localization and navigation.
• The internal temperature mechanism successfully balanced exploration and exploitation during the search.

Conclusions:

• The developed scheme provides a viable model for understanding animal navigation strategies in complex environments.
• The algorithm offers potential applications for robotic systems in challenging, dynamic media, reducing reliance on odometry.
• This approach is beneficial for problems requiring active control over the exploration-exploitation trade-off.