Spatio-Temporally Smooth Local Mapping and State Estimation inside Generalized Cylinders with Micro Aerial Vehicles
Published in IEEE Robotics and Automation Letters (RA-L), also presented at ICRA, 2018
This paper introduces a novel approach for real-time state estimation and egocentric local mapping using micro aerial vehicles (MAVs) operating inside non-cylindrical tunnels, modeled as parameterized piecewise-smooth generalized cylinders (PSGCs). The system leverages a 3D LiDAR and an IMU, employing a constrained Unscented Kalman Filter (UKF) to perform robust localization along changing cross-sections and nonlinear surfaces.
The method adapts to variable tunnel geometries and outperforms previous work in mapping quality and robustness, enabling fully autonomous inspection flights with minimal operator intervention. The approach was validated through over 20 real-world deployments, including field experiments in penstocks at the Center Hill Dam, TN.
Recommended citation: Özaslan, T., Loianno, G., Keller, J., Taylor, C. J., & Kumar, V. (2018). "Spatio-Temporally Smooth Local Mapping and State Estimation inside Generalized Cylinders with Micro Aerial Vehicles." IEEE Robotics and Automation Letters, 3(3), 1755–1762.
Download Paper
Related people: Tolga Ozaslan