Papers & Talks

Proposes a geometric clustering approach for distributed fault detection and identification in range-based cooperative localization, enabling robust multi-agent localization without relying on fixed reference (anchor) nodes.

Presents an optimization-based motion planning approach for vector field following in dynamic environments, enabling UAVs to track smooth trajectories while reacting to changing obstacles and environmental conditions.

Experimental analysis of hover constraints in quadcopter UAVs and how they can be exploited as pseudo-measurements to improve localization accuracy in cooperative navigation scenarios.

Analyzes the unscented transform (UT) as a nonlinear estimation tool for cooperative localization using only inter-vehicle ranging measurements, evaluating accuracy and consistency in GNSS-denied conditions.

A cooperative localization framework for a UAV–UGV team in GNSS-denied subterranean environments. Integrates Factor Graph Optimization for UAV localization and SLAM for UGV navigation, validated through simulation and real-world tunnel experiments.

A LiDAR-based UAV position estimation method using PointPillars deep learning architecture. A dataset of 7,000+ annotated LiDAR scans was created, achieving a 30% improvement in localization accuracy over traditional point cloud clustering methods.