First EUSOME Publication Accepted
We are pleased to announce that the first scientific publication resulting from work conducted within the EUSOME project has been accepted for publication.
The article, titled “An Onboard UAV Multi-task System for Trajectory Prediction and State Estimation Employing Transformer- and Reservoir-based Networks” by N. Souli, P. Katzanis, P. Kardaras, Y. Grigoriou, S. Stavrinides, P. Kolios, and G. Ellinas, has been published in the Journal on Autonomous Transportation Systems.
🔗 Read the publication
The study introduces a real-time onboard multi-task learning system for UAVs that integrates Transformer neural networks with reservoir computing architectures. This hybrid approach enables robust and accurate UAV state identification and trajectory prediction by leveraging sequential sensor data collected during real-world UAV operations. The system is trained on multimodal measurements and has been validated in field experiments, demonstrating high accuracy and practical applicability in diverse operational conditions.
This publication marks an important dissemination milestone for the project.
A new Publications section has also been added to the EUSOME website, where all future project-related publications will be listed.
