Meet the AUTOASSESS project coordinator: Technical University of Denmark (DTU)

May 23, 2024

AUTOASSESS, the project that aims to remove human surveyors from dangerous enclosed areas such as ballast tanks or cargo holds, has a consortium formed by 16 members from 9 countries.

The first Meet the Partner article is about the Technical University of Denmark (DTU), the project’s coordinator, and it allows a better understanding of this partner’s role and objectives in the project.

DTU aspires to become an international center for technological development, achieved through education, cutting-edge research, innovation, and scientific consultancy. Through technical and natural sciences, DTU generates value for society and works towards a sustainable future by developing technology that serves people’s needs.

Foremost, as the coordinator of the AUTOASSESS, DTU is managing the project’s implementation and overviewing the development. While being directly involved in 9 out of 10 Work Packages (WP), DTU is the WP Lead of WP 5 where it will contribute to the visual defect detection and semantic segmentation of the geometric structure. The expertise of the DTU project team is also in perception-based viewpoint optimization (WP4), as well as fusion of detection data into the digital twin (WP6).

The DTU project team involved has a history of research in autonomous robotic inspection, with INSPECTRONE being the precursor project to AUTOASSESS:

Associate Professor Evangelos Boukas: Next to coordinating the AUTOASSESS project and leading the DTU project team, associate professor Evangelos Boukas is the head of the Autonomous Systems MSc at DTU. He brings along significant experience from both national and European projects in field robotics, space robotics, inspection robotics and industrial robotics. His research interests in AUTOASSESS are Autonomous Robot Operation and Machine Learning for Intelligent Robot Perception.
Post Doctoral Researcher Rasmus Eckholdt Andersen: Rasmus is the WP Lead for WP 5 and specializes in Machine Learning-based perception for Inspection. His experience in ambiguous classification adds value to the AUTOASSESS project by discerning between essential structural components and potential hazardous defects.
PhD student Susan Basnet: Susan is a PhD student at DTU Electro working on the topic “Aerial Robot Design and Control System for Inspection of Confined Spaces” under the supervision of Associate Professor Evangelos Boukas. His specialization mainly focuses on control system design for the non-linear systems. Being an Aerospace Engineering graduate, he has also worked in designing aerial vehicles and implementing guidance and navigation strategies on them.
PhD student Dimitrios Syrrafos: Dimitrios is a PhD student at DTU Electro working on the topic “Active Semantic Segmentation and Probabilistic Learning for Inspection of Confined Spaces using Aerial Robots” under the supervision of Associate Professor Evangelos Boukas. With a master’s degree in Data Science & Machine Learning, Dimitrios is an electrical and computer engineer who conducted his master’s thesis on geometric deep learning. Dimitrios pushes the envelope in machine learning for non-deterministic inference in ambiguous segmentation and defect detection problems.
Project Manager Melanie Brunhofer: As project manager of the AUTOASSESS project, Melanie can rely on experience from previous big scale European funded projects (FP7, H2020 and Horizon Europe) in different thematic areas.

At DTU Electro the immense possibilities of electrical and photonics engineering are combined.

DTU sees AUTOASSESS as the cornerstone of autonomous robotics inspection for confined, GNSS-denied, critical infrastructures. Specifically in the maritime domain, AUTOASSESS will provide quick, accurate and extremely safe marine vessel inspections. By employing autonomous superhuman drones, surveyors are kept out of harm’s way and Millions of Euros are saved for the industry.