Paulo Abelha

University of Aberdeen, King's College Hospital

Papers

Total Citations

H-Index

About

Paulo Abelha’s research lies at the intersection of robotics, computer vision, and artificial intelligence, with a core focus on enabling robots to understand and manipulate tools in dynamic, unstructured environments. His major contribution is a model-based framework that allows robots to recognize and reason about tool affordances—the functional properties that make an object suitable for a task—using 3D vision data. In his most cited work, “A model-based approach to finding substitute tools in 3D vision data” (27 citations), Abelha demonstrated how a robot can leverage prior knowledge of known tools (e.g., a hammer) to identify and use unfamiliar objects as functional substitutes when the original tool is unavailable. This was extended in “Learning how a tool affords by simulating 3D models from the web” (26 citations), where he showed that robots can learn affordances by simulating 3D models sourced online, dramatically reducing the need for real-world training data. His work has been recognized for its practical impact on service robotics, particularly in domestic settings like kitchens, where adaptability is critical. Abelha’s research not only advances autonomous manipulation but also provides a scalable pathway for robots to improvise and operate beyond controlled environments.