Intermittent Information-Driven Multi-Agent Area-Restricted Search

Area of Science:

• Robotics
• Autonomous Systems
• Search Theory

Background:

• Area-restricted search problems involve locating targets within defined boundaries.
• Autonomous mobile sensing platforms (agents) are increasingly used for surveillance.
• Underwater environments present unique challenges for sensing and communication.

Purpose of the Study:

• To develop a decentralized, information-driven search strategy for a team of autonomous agents.
• To address the challenges of range-dependent detection probability and false alarms in sensing.
• To enable efficient underwater surveillance using amphibious drones.

Main Methods:

• A novel intermittent search strategy alternating between ballistic (fast displacement) and diffusive (slow sensing) phases.
• Decentralized computation for estimation and motion control, relying on local data.
• Neighbor-only data exchange for agent coordination, eliminating the need for global network knowledge.

Main Results:

• The proposed strategy enables coordinated multi-agent search in a decentralized manner.
• Effective target detection is achieved despite range-dependent sensing probabilities and false alarms.
• The approach facilitates efficient underwater surveillance with amphibious drone swarms.

Conclusions:

• The intermittent information-driven search strategy is effective for decentralized multi-agent systems.
• This method offers a robust solution for underwater surveillance and area-restricted search problems.
• Local computations and neighbor-only communication ensure scalability and adaptability.