Duy Nguyen-Tuong

About

Duy Nguyen-Tuong is a leading researcher at the intersection of machine learning and robotics, with a particular focus on model learning for robot dynamics and real-time control. His work has fundamentally advanced how robots acquire accurate models of their own mechanics, enabling more precise, energy-efficient, and compliant control systems. Among his most influential contributions is his pioneering development of Local Gaussian Process Regression (LGP), a technique that dramatically accelerated standard Gaussian Process methods to meet the demanding speed requirements of real-time robot control. This line of work, explored across multiple highly cited publications between 2008 and 2011, demonstrated that machine learning could serve as a practical alternative to analytically derived rigid-body models when unmodeled nonlinearities limit their accuracy. His 2011 survey, "Model Learning for Robot Control," has become a definitive reference in the field with nearly 500 citations, while his LGP papers collectively represent a foundational body of work reshaping robot learning methodology. Nguyen-Tuong also made significant strides in integrating prior physical knowledge with data-driven learning, reducing sample complexity without sacrificing accuracy. His research on real-time trajectory planning further extended his impact into dynamic manipulation tasks. Collectively, his work has garnered over 1,700 citations, cementing his reputation as a key architect of modern learning-based robot control.

Research Focus

Key Achievements

Top Papers

1. 1
   Model learning for robot control: a survey

   499 citations · 2011
2. 2
   Model Learning with Local Gaussian Process Regression

   322 citations · 2009
3. 3
   Using model knowledge for learning inverse dynamics

   176 citations · 2010
4. 4
   Computed torque control with nonparametric regression models

   151 citations · 2008
5. 5
   Local Gaussian process regression for real-time model-based robot control

   144 citations · 2008
6. 6
   Local Gaussian Process Regression for Real Time Online Model Learning

   121 citations · 2008
7. 7
   Trajectory planning for optimal robot catching in real-time

   117 citations · 2011
8. 8
   Local Gaussian process regression for real time online model learning and control

   88 citations · 2008
9. 9
   Incremental online sparsification for model learning in real-time robot control

   73 citations · 2011
10. 10
    Learning Inverse Dynamics: A Comparison

    65 citations · 2008

Key Collaborators

Contact & Links