Abstract
The small footprint of tiny multirotor vehicles is advantageous for accessing tight spaces, but their limited payload and endurance impact the ability to carry powerful sensory and computing units for navigation. This article reports an aerodynamics-based strategy for a ducted rotorcraft to avoid wall collisions and explore unknown environments. The vehicle uses the minimal sensing system conventionally conceived only for hovering. The framework leverages the duct-strengthened interaction between the propeller wake and vertical surfaces. When incorporated with the flight dynamics, the derived momentum-theory-based model allows the robot to estimate the obstacle’s distance and direction without range sensors or vision. To this end, we devised a flight controller and reactive navigation methods for the robot to fly safely in unexplored environments. Flight experiments validated the detection and collision avoidance ability. The robot successfully identified and followed the wall contour to negotiate a staircase and evaded detected obstacles in proof-of-concept flights.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have