Autonomous Field Robotics: Coverage, Exploration, and Mapping

ABSTRACT

My research is focused on strategies for making robots operate autonomously in the real world even in the presence of adversarial conditions: low visibility, wind, currents, dynamic obstacles, confined environments, etc. The marine environment, below and above water, is the least studied on Earth and one of the most challenging. While most of research in marine robotics is focused on open areas where the robot operates through free space, my work is focused on the coastal and inland areas where the vehicle needs to maneuver around obstacles, the environment dynamics present the most challenges, and autonomy is a key requirement. This talk will start by addressing multi-robot efficient coverage of large bodies of water, complete or partial, and single-robot coverage of confined water bodies such as rivers. On the topic of exploration, I will cover multi-robot exploration of shipwrecks and single-robot exploration of underwater caves. Regarding mapping, I will discuss multi-sensor fusion for mapping shipwrecks, underwater caves, and coral reefs. Results on machine learning approaches for semantic mapping including coral classification and cave image segmentation will be discussed. The work that I will present has a strong algorithmic flavour, while it is validated in the real world. Experimental results from several testing campaigns will be presented.

BIOGRAPHY

Ioannis Rekleitis is an Associate Professor at the Department of Mechanical Engineering, University of Delaware (2025-present), and an Adjunct Professor at the School of Computer Science, McGill University. He was an Assistant/Associate/Full Professor at the Computer Science and Engineering Department, University of South Carolina (2014-2024). During 2004-2007 he was a visiting fellow at the Canadian Space Agency working on Planetary exploration and On-Orbit-Servicing of Satellites. During 2004 he was at McGill University as a Research Associate in the Centre for Intelligent Machines with Professor Gregory Dudek in the Mobile Robotics Lab (MRL). Between 2002 and 2003, he was a Postdoctoral Fellow at the Carnegie Mellon University in the Sensor Based Planning Lab with Professor Howie Choset. He was granted his Ph.D. from the School of Computer Science, McGill University, Montreal, Quebec, Canada in 2002 under the supervision of Professors Gregory Dudek and Evangelos Milios. Thesis title: "Cooperative Localization and Multi-Robot Exploration". He completed his M.Sc. in McGill University in the field of Computer Vision in 1995. He was granted his B.Sc. in 1991 from the Department of Informatics, University of Athens, Greece. His Research has focused on mobile robotics and in particular in the area of autonomous field robotics systems, developing multi-robot cooperative localization, mapping, exploration and coverage algorithms for the marine environment. His interests extend to computer vision, learning, and sensor networks.