
Robotics
About Domain
<p>We gather several competences that allow us to develop advanced robotic solutions capable of operating in challenging environments, with autonomy, precision and safety. Some of them include:</p>

Autonomous Navigation

Marine, Agro and Industrial Robotics

Localisation and Mapping

Human-Robot Collaboration
Challenges
<p>The work developed by our Robotics scientific domain is at the forefront of developing real multi-domain robotics. It combines intelligence, autonomy, and usefulness seamlessly across various uses on land, in the air, on the water, and underwater. We lead the way in developing new and creative scientific methods that connect different areas, resulting in a cohesive foundation for robotic systems. Our main challenges are:</p>
Main achievements
<p>Our research in Robotics drives the development of advanced systems capable of operating with autonomy, precision, and safety in demanding and dynamic environments. Our key achievements combine environmental perception, human-robot collaboration, and intelligent navigation, with applications in industrial, marine, and agricultural contexts.</p>

Advanced Perception for Environmental Mapping
<p>We developed successive generations of perception systems and data processing algorithms that allow us to map complex environments in 3D with high precision, even at great underwater depths. These systems combine high-resolution imaging with miniaturisation capabilities and have been integrated into various underwater robots. Learn more <a href="https://adgeo.copernicus.org/articles/62/1/2023/" target="_blank">here</a>, <a href="https://www.sciencedirect.com/science/article/abs/pii/S156625351830366X" target="_blank">here</a>, <a href="https://link.springer.com/article/10.1007/s10846-017-0689-0" target="_blank">here</a>, and <a href="https://ieeexplore.ieee.org/document/9024043" target="_blank">here</a>.</p>

Human-Robot Collaboration for Industrial Assembly
<p>We developed a cognitive system for collaborative robots (cobots) in engine assembly operations. By combining computer vision and deep learning, the system interprets the gestures and actions of the human operator and autonomously adjusts the robot's assembly plan. The model achieved an accuracy of 96.65% in interpreting human actions, demonstrating great potential for collaborative industrial environments. Learn more <a href="https://www.researchgate.net/publication/371063285_Sensor_Placement_Optimization_using_Random_Sample_Consensus_for_Best_Views_Estimation" target="_blank">here</a> and <a href="https://www.sciencedirect.com/science/article/abs/pii/S0736584522001314" target="_blank">here</a>.</p>

Semantic Mapping and Localization for Mobile Robots
<p>We created NAVIBOX, an innovative mapping and localisation solution that integrates metric, topological, and semantic information for mobile robots in outdoor environments. It includes advanced path planning algorithms, obstacle avoidance, and a mission supervisor that translates agronomic maps into autonomous robotic actions. NAVIBOX was successfully tested on new agricultural robots developed by INESC TEC. Publications: <a href="https://onlinelibrary.wiley.com/doi/abs/10.1002/rob.22140" target="_blank">here</a> and <a href="https://www.sciencedirect.com/science/article/abs/pii/S0921889021000105" target="_blank">here</a>.</p>