Minimal navigation solution for a swarm of tiny flying robots to explore an unknown environment

Area of Science:

• Robotics
• Artificial Intelligence
• Swarm Intelligence

Background:

• Tiny flying robots offer potential for indoor exploration but lack navigation strategies.
• Existing solutions like SLAM are resource-intensive for small robots.
• Autonomous navigation requires robots to perform their own positioning without external infrastructure.

Purpose of the Study:

• To present a minimal navigation solution for swarms of tiny flying robots.
• To enable autonomous exploration and return to the departure point.
• To demonstrate effective swarm behavior for coverage and collision avoidance.

Main Methods:

• Developed the swarm gradient bug algorithm (SGBA) for navigation.
• Utilized visual odometry and wall-following for obstacle avoidance.
• Implemented inter-robot communication for collision avoidance and efficiency.
• Employed a gradient search toward a home beacon for return navigation.

Main Results:

• Demonstrated successful exploration of a real-world environment by a swarm of 33-g quadrotors.
• Maximized coverage by directing robots in diverse directions from the departure point.
• Proved the algorithm's effectiveness in a search-and-rescue proof-of-concept mission.

Conclusions:

• SGBA provides a minimal, effective navigation strategy for robot swarms in unknown indoor environments.
• The algorithm is generalizable to different robot types.
• This work lays the foundation for complex missions using robot swarms, including search and rescue.