Artificial pheromone for path selection by a foraging swarm of robots

Area of Science:

• Robotics
• Artificial Intelligence
• Swarm Intelligence

Background:

• Foraging robots navigate environments for search and retrieval tasks, often creating paths for efficiency.
• Swarm-based path selection is crucial for optimizing foraging behavior and resource allocation.
• Existing methods may lack adaptability to dynamic resource availability or robot task switching.

Purpose of the Study:

• To implement and validate a path selection mechanism for foraging robot swarms.
• To enhance swarm foraging efficiency by enabling robots to focus on the most profitable resources.
• To allow robots to dynamically switch tasks and adapt to changing environmental conditions.

Main Methods:

• Implementation of virtual ants transmitting local messages to simulate artificial pheromone trails.
• Robots use pheromone concentration to determine path selection and maintenance.
• Mathematical modeling and experimental validation with a swarm of 20 real robots.

Main Results:

• The proposed mechanism successfully favors the selection of the closest resource.
• The system demonstrates the ability to select a new path when a resource becomes unavailable.
• The swarm can adapt to select newly detected, more profitable resources.

Conclusions:

• The virtual ant-based pheromone system effectively enables path selection in robot swarms.
• The mechanism is robust, adapting to resource unavailability and discovering better resources.
• The simplicity of the robot messages and behaviors makes it suitable for microrobot swarms.