Research / 02
My doctoral research explores semantic simultaneous localisation and mapping (SLAM), situational awareness, and collaborative missions for unmanned aerial vehicles (UAVs).
SnT, University of Luxembourg, Kirchberg · Advisor: Prof. Dr Holger Voos · Nov 2025 - present.
The aim is to move beyond purely geometric mapping so that a UAV can reason about objects, places, and the relations between them while it localises and maps. Current work covers C++ and Python implementation with OpenVINS-based estimation running on embedded Linux, using Raspberry Pi hardware with camera and IMU sensing.
The aim is an onboard architecture that turns sensing into an understanding of the situation in order to support navigation decisions. Current work covers ROS 2 and PX4 integration, Docker-based tooling, Kalibr camera and IMU calibration, and testing across simulation, laboratory, hardware, and flight settings.
The longer-term direction is for teams of UAVs to conduct collaborative missions using shared semantic maps, resource-efficient navigation, and mission-aware coordination.
Supporting simulation background is collected on the projects page: the N-Body Simulator rigid-body and gravity framework with its attitude-estimation harness, the ArvSimulator atmospheric trajectory simulation, and the lunar simulator rover autonomy stack.
Student-team systems work is described under CubeSat projects, including IDeT-Sat, with its paper on the publications page.
Photographs from the SaCoGraph presentation and space-engineering programmes are collected alongside their training entries.