Max Robotics — Global Service Robot Directory + US Certification

About

Luigi Nardi is a researcher whose work spans simultaneous localization and mapping (SLAM), computer vision, embedded systems, and robot learning. He has made significant contributions to the development of efficient volumetric SLAM frameworks, most notably through his work on octree-based representations supporting both signed-distance and occupancy mapping, which has garnered 114 citations and remains a cornerstone reference in dense 3D scene reconstruction. His development of SLAMBench2 — a rigorous multi-objective benchmarking framework for visual SLAM — has provided the robotics and augmented reality communities with a much-needed standardized evaluation methodology, accumulating nearly 90 citations across publications. Nardi has also contributed influential survey-level work navigating the computational landscape of real-time localization and mapping, addressing the pressing challenge of deploying high-performance vision systems on power-constrained embedded platforms. More recently, his research has evolved toward robot skill learning, combining reinforcement learning, Bayesian optimization, and behavior trees to enable efficient, safe, and user-informed acquisition of complex industrial manipulation tasks. His work on adaptive Bayesian optimization in nested subspaces further demonstrates his growing expertise in sample-efficient optimization for high-dimensional problems. Across his career, Nardi has established himself as a versatile contributor bridging computer vision, robotics, and machine learning.