Research

We work on various mobile robot implementations together with perception, planning, and control algorithms to enable various autonomous robotic operations. Tasks such as system integration, architecture and real-time processing are also key aspects of our work to ensure seamless operation, efficient resource utilization, and timely decision-making in dynamic environments. Open source or proprietary operating systems are selected depending on the requirements and specifications of the required robotic tasks and limitations. Extensive simulations are also performed based on several simulators.

Autonomous vehicles (Level 3+) and autonomous driving present unique challenges, from navigating dense urban environments to operating in extreme weather and complex traffic conditions. Our research focuses on advancing their perception and localization systems together with their planning algorithms. We explore efficient sensor fusion techniques to enhance robustness and accuracy in object detection, recognition and tracking. Our work also included the development of the first Hellenic (Greek) Autonomous Vehicle based on a commercial electric vehicle.

Deep learning, a subset of machine learning, utilizes algorithms designed to capture and model complex abstractions within data. In our research, we leverage deep learning techniques to address various mobile robot navigation challenges, including path estimation, perception, end-to-end navigation, and multi objective strategies.

At the forefront of robotics research, we focus on advancing multi-sensor perception systems to enable robust and reliable robot autonomy in complex environments. We integrate data from LiDAR, RGB(D) cameras, inertial measurement units (IMUs), Radar and GNSS sensors to achieve high-resolution environmental understanding and real-time decision-making. Our current projects include developing novel sensor fusion algorithms that combine traditional and deep learning-based approaches to enhance accuracy in object detection, semantic segmentation, and 3D mapping.

Our research focuses on the development of intelligent systems that foster social engagement and support diverse learning needs, including those in special education. We utilize humanoid and non-humanoid robots equipped with advanced speech recognition, emotion detection, and adaptive learning algorithms to create personalized, interactive learning experiences. Our research explores the role of robotics in promoting collaboration, problem-solving, and emotional intelligence among students, while also addressing the unique challenges faced in special education.