Javier Segovia‐Aguas

Institut de Robòtica i Informàtica Industrial, Pompeu Fabra University

Papers

Total Citations

H-Index

About

Javier Segovia‐Aguas is a researcher at the intersection of artificial intelligence, robotics, and cognitive science, with a primary focus on socially assistive robotics and automated planning. His work centers on enabling robots to autonomously learn and teach cognitive training exercises, particularly for elderly or rehabilitation patients. In his most cited work, "Natural Teaching of Robot-Assisted Rearranging Exercises for Cognitive Training" (2019, 14 citations), Segovia‐Aguas developed methods for robots to naturally instruct users in cognitive tasks, bridging human-robot interaction with therapeutic applications. A key theoretical contribution is his "Planning with Partially Specified Behaviors" (PPSB) framework (2016, 5 citations), which elegantly combines reinforcement learning with classical planning to solve sequential decision problems—demonstrating that these two often-separate paradigms can complement each other effectively. More recently, his 2021 paper on "Automatic Learning of Cognitive Exercises" (4 citations) introduced a novel approach for teaching assistive robots new board exercises by learning action models from demonstrations, using Boolean predicates and existential quantifiers. This work has significant implications for personalized, adaptive robotic therapy, allowing caregivers to easily program new cognitive activities without technical expertise.