Concetto Spampinato

Papers

Total Citations

H-Index

About

Concetto Spampinato is a leading researcher in robotics and artificial intelligence, with a primary focus on autonomous navigation in unstructured outdoor environments. His work centers on developing self-supervised learning techniques that enable robots to predict terrain traversability from RGB images without requiring costly manual annotations—a critical advancement for field robotics. By combining unsupervised domain adaptation with synthetic data, Spampinato has created robust models that transfer knowledge from simulation to real-world deployment, significantly reducing the data bottleneck in robot learning. His most-cited paper (2024) demonstrates this approach, achieving 7 citations for its practical impact on safe robot locomotion. Beyond navigation, Spampinato has contributed to agent-based software architectures for autonomous systems, as shown in his early work on the Eurobot 2006 competition, where he applied multi-agent control to a golf-playing robot. He also developed MIDGARD, a specialized robot navigation simulator for outdoor unstructured terrains, providing a vital tool for the research community. Spampinato’s work bridges simulation and reality, advancing the frontier of autonomous exploration in challenging environments.