Max Robotics — Global Service Robot Directory + US Certification

About

Hao Fan is a leading researcher in the field of autonomous robotics, with a primary focus on Simultaneous Localization and Mapping (SLAM) and visual-inertial navigation. His work addresses critical challenges in enabling robust perception for robots operating in complex and dynamic environments, from underwater exploration to aerial systems. Fan’s most influential contribution is his pioneering work on loop closure detection for visual SLAM, where he introduced the use of PCANet features—an unsupervised deep learning approach—to significantly improve map consistency and reduce accumulated error. This foundational paper has garnered 44 citations, underscoring its impact on the SLAM community. He has further advanced the field by developing a marker-based method for visual-inertial initialization, which is crucial for accurate and reliable sensor fusion, and by tackling the problem of dynamic scene understanding through semantic SLAM that examines moving consistency. Looking toward the future, Fan is pushing the boundaries of real-time autonomous exploration with limited field-of-view sensors, a key enabler for underwater and aerial vehicles. His recent work on learning semantic-aware point-line features promises to enhance localization and 3D reconstruction in challenging environments, marking him as a rising innovator in next-generation robotic perception.