Pedro Gusmao

Papers

Total Citations

H-Index

About

Pedro Gusmão is a leading researcher at the intersection of robotics, autonomous navigation, and sensor fusion, with a particular focus on robust perception in challenging environments. His most influential work introduces a groundbreaking Graph-Based Thermal–Inertial SLAM system that leverages probabilistic neural networks to overcome the limitations of traditional vision-based localization. By harnessing thermal imaging, Gusmão’s approach enables reliable simultaneous localization and mapping even in adverse visibility conditions—such as darkness, smoke, or airborne particulates—where conventional cameras fail. This innovation, published in 2021 and already garnering 46 citations, represents a significant leap forward for field robotics in disaster response, underground exploration, and industrial inspection. Beyond this flagship contribution, Gusmão’s broader research advances the integration of deep learning with state estimation, pushing the boundaries of how autonomous systems perceive and navigate the physical world. His work is widely recognized for its practical impact, bridging the gap between theoretical probabilistic methods and real-world deployment in GPS-denied or visually degraded settings. For students and researchers, Gusmão exemplifies how combining domain-specific sensing with modern machine learning can solve enduring challenges in robotic autonomy.