'Infotaxis' as a strategy for searching without gradients

Area of Science:

• Biology
• Robotics
• Physics

Background:

• Chemotactic bacteria use local gradients for navigation.
• Macroscopic searchers face challenges due to intermittent cues in flowing media.
• Animals detect odors sporadically as patches move with currents.

Purpose of the Study:

• To propose a novel search algorithm, infotaxis, for navigating with sparse information.
• To provide a strategy for macroscopic searchers in environments with fragmented cues.
• To develop a method for maximizing information gain during a search.

Main Methods:

• Computational modeling of odor plume propagation.
• Analysis of experimental data from mixing flows.
• Development of the infotaxis algorithm based on information gain maximization.

Main Results:

• Infotaxis effectively guides searchers in environments with intermittent cues.
• Simulated and experimental results validate the infotaxis strategy.
• Infotactic trajectories mimic natural behaviors like zigzagging and casting.

Conclusions:

• Infotaxis offers an efficient search strategy for intermittent environmental cues.
• The algorithm is applicable to olfactory robot design.
• The principles of infotaxis can be generalized to various sparse information search problems.