Decentralized Processing for Distance-Based Formation Control

Multi-agent System (Swarm)

• A coordinated group of physical agents operating in a shared environment, tasked with the execution of a collective goal or individual goals.

• Complex task execution, scalability, robustness, and adaptability.

• Challenges: Coordination, sensing, communication, collision avoidance. etc.

• Applications: surveillance, cooperative transportation, disaster response, space exploration. etc.

TIGER Square Multi-agent Robotic Testbed

• Three major test components : A square test arena (TIGER Square), the Control Station, and the TIGERBots.

• Control station: A control computer, a Wi-Fi router, an overhead localization camera.

• A fleet of 10 Tiny Intelligently Grouped Experimental Robots (TIGERBots).

  • Two modular circuit boards: A main board that handles the locomotion, communication and processing. An Infrared sensor board for decentralized sensing. 

  • Nonholonomic, differentially driven with two stepper motors. 3.5 cm/s heading speed, 45 deg/s angular speed.

  • Global localization performed by detecting unique fiducial marker tag on top each TIGERBots.

Operation

• 4 operating modes based on Centralized or Decentralized sensing, processing, and communication methods.

2. Formation Manuevering

Simultaneously acquire a formation, and maneuver cohesively as a virtual rigid body along a predefined trajectory.