Learning-Based Perception and Control for Intelligent Decision-Making

Antonio Loquercio is an Assistant Professor at the University of Pennsylvania. He earned his PhD and M.Sc. from UZH and ETH Zurich in 2021 and 2017. His research lies at the intersection of learning-based robotics, computer vision, and machine learning, with pioneering contributions to transferring sensorimotor control from simulation to the real world.

He has received several prestigious honors, including the ETH Medal for Outstanding Master Thesis (2017), the Best System Paper Award at the Conference on Robot Learning (2018), honorable mentions for Best Paper at RSS (2020) and T-RO (2020), and the Georges Giralt PhD Award — Europe’s top recognition for doctoral research in robotics. His work on superhuman drone racing, featured on the cover of Nature, demonstrated the first autonomous robot to outperform a human champion.

At the heart of Loquercio’s research is a fundamental question: What is the true purpose of perception in autonomous systems? While traditional approaches aim to recreate detailed replicas of the world, his work focuses on developing perceptual models that are sufficiently accurate to support effective decision-making and control. By quantifying how model fidelity impacts system performance, his research seeks to bridge the gap between perception and action.

This paradigm is not only advancing the capabilities of intelligent robots but also paving the way for controlling complex, dynamic systems that remain challenging to model—such as large-scale wind farms or power grids. Ultimately, these advances could drive more efficient, adaptive, and intelligent control across a wide range of real-world applications.