Max Robotics — Global Service Robot Directory + US Certification

About

Pablo Guerrero’s research lies at the intersection of probabilistic robotics, active vision, and autonomous decision-making, with a strong emphasis on enabling robots to perceive and act reliably in dynamic, real-world environments. His work is deeply rooted in the RoboCup domain, where he contributed foundational methods for legged robot vision and control. Guerrero’s most influential paper, “Evolving Visual Object Recognition for Legged Robots” (16 citations), pioneered evolutionary approaches to robust perception, while “Probabilistic Decision Making in Robot Soccer” (15 citations) advanced the use of Bayesian reasoning for strategic action under uncertainty. He also made key contributions to context-dependent color segmentation for Aibo robots, addressing the persistent challenge of variable lighting in vision systems. Guerrero’s work on explicitly task-oriented probabilistic active vision, which focuses on reducing uncertainty most relevant to the robot’s current goal, represents a significant conceptual advance in active perception. His research on cooperative global tracking with multiple sensors and Bayesian spatiotemporal context integration further demonstrates his commitment to building robust, scalable perception systems. As a member of the UChile Kiltros team, Guerrero helped bridge developments across RoboCup leagues, fostering cross-platform innovation. With a career spanning over a decade, his work has shaped how mobile robots learn to see, decide, and act in complex, uncertain environments.