Inspection of Pole-Like Structures Using a Visual-Inertial Aided VTOL Platform with Shared Autonomy

Area of Science:

• Robotics and Automation
• Computer Vision
• Aerospace Engineering

Background:

• Vertical infrastructure inspection is challenging, posing risks and high costs.
• Micro vertical take-off and landing (VTOL) unmanned aerial vehicles (UAVs) offer a safer, cost-effective alternative.
• Existing systems face challenges in close-proximity flight, disturbance rejection, and remote operation.

Purpose of the Study:

• To develop and evaluate a VTOL UAV system for autonomous vertical infrastructure inspection.
• To address challenges in maintaining safe stand-off distances and rejecting environmental disturbances.
• To compare the performance of Position-Based Visual Servoing (PBVS) and Image-Based Visual Servoing (IBVS) for this application.

Main Methods:

• Utilized a VTOL UAV equipped with monocular vision, inertia, and sonar sensors.
• Implemented two visual servoing approaches: PBVS with an Extended Kalman Filter (EKF) and estimator-free IBVS.
• Conducted simulations and indoor/outdoor flight experiments under various conditions (day/night).

Main Results:

• Demonstrated successful close-quarters operation, maintaining safe stand-off distances from vertical poles.
• Showcased robustness against environmental disturbances like wind gusts.
• Validated the system's capability for inspection and navigation around vertical structures in GPS-impaired environments.

Conclusions:

• The developed VTOL UAV system effectively performs vertical infrastructure inspection with shared autonomy.
• Both PBVS and IBVS approaches show promise, with performance varying in accuracy, robustness, and computational cost.
• The system offers a viable solution for hazardous or time-consuming inspection tasks.