Mengda Geng

University of Science and Technology Beijing

Papers

Total Citations

H-Index

About

Mengda Geng is a researcher focused on advancing human-robot collaboration, particularly in physical co-manipulation tasks. His key research areas include human motion prediction, shared control, and adaptive robot behavior in dynamic environments. Geng’s major contribution is the development of a long short-term human motion prediction method using LSTM networks for human-robot co-carrying. This work enables robots to anticipate a human partner’s motion target, allowing the robot to proactively lead the task rather than merely react—enhancing efficiency and natural interaction. With 4 citations, this foundational paper addresses a critical challenge in physical human-robot interaction: predicting long-term human intent to achieve seamless coordination. Geng’s approach stands out for its focus on practical, real-time applications in collaborative manufacturing and assistive robotics. His work has been presented at leading robotics venues, and he continues to explore how predictive models can make human-robot teams more intuitive and productive. For students and researchers, Geng’s research offers a compelling entry point into the intersection of deep learning and physical human-robot interaction.