A self-rotating, single-actuated UAV with extended sensor field of view for autonomous navigation

Area of Science:

• Robotics
• Aerospace Engineering
• Computer Vision

Background:

• Uncrewed Aerial Vehicles (UAVs) typically have limited perception due to small sensor fields of view (FoV).
• Extending the FoV is crucial for enhancing UAV navigation, exploration, and obstacle detection capabilities.
• Leveraging inherent self-rotation from motor counter-torque offers a power-efficient method to expand sensor FoV.

Purpose of the Study:

• To develop an agile, self-rotating UAV (PULSAR) capable of panoramic sensing using a single actuated motor.
• To demonstrate reduced power consumption compared to traditional quadrotors.
• To enable autonomous navigation and obstacle detection in unknown environments.

Main Methods:

• Designed PULSAR, an ultra-underactuated aerial robot controlled by a single motor for thrust and moment generation.
• Integrated an onboard LiDAR sensor for 3D environment perception.
• Implemented control strategies to manage coupled dynamics during self-rotation for navigation.

Main Results:

• PULSAR achieved autonomous navigation and obstacle avoidance with an extended FoV through self-rotation.
• Demonstrated a 26.7% reduction in power consumption compared to a benchmark quadrotor.
• Maintained agility while performing environment exploration and dynamic obstacle avoidance.

Conclusions:

• Self-rotating UAVs like PULSAR can significantly enhance perception capability and task efficiency.
• Single-actuator control is viable for agile flight and reduced power usage.
• This approach offers a promising direction for safer and more capable autonomous aerial systems.