Computational Scientist & Software Engineer

Mujin Inc., Tokyo, Japan

PhD of Computational Science and Engineering

McMaster University, Canada



Research interest: 

  • Robot kinematics: inverse kinematics, system identification
  • Numerical analysis, automatic differentiation, optimization algorithms
  • Ordinary differential equations, differential-algebraic equation systems
  • Scientific computing and programming in Matlab/C++/Python


Guangning Tan (谭广宁), was born and raised in Guangzhou, China. He received in 2010 his Bachelor degree in Communication and Electrical Engineering at Sun Yat-sen University, Guangzhou, and in 2012 and 2016 his M.Eng and PhD respectively in Computational Science & Engineering at McMaster University, Canada. While at McMaster, he was the main author of the DAESA package, a Matlab tool for structural analysis of differential-algebraic equation systems.

After finishing post-doc at MIT, Tan joined Mujin Inc. (Tokyo, Japan) as Computational Scientist and Software Engineer. He has succeeded in solving one of the most difficult computational problems in robot kinematics automation. He authored IKFastCpp, a C++ package that solves all kinds of inverse kinematics (IK) problems of all industrial robots. Within IKFastCpp he has developed innovative mathematical theories, numerical algorithms, and efficient modern C++ code. Compared with IKFastCpp's predecessor originated by Mujin's CTO Dr. Rosen Diankov, IKFastCpp can generate much more reliable and versatile solution code in any programming language. The Mujin company uses these inverse kinematics solvers in numerous robotics automation applications, such as motion planning, and kinematics calibration in robot system identification.


Selected Publications

DAESA Theory

DAESA: A Matlab Tool for Structural Analysis of Differential-Algebraic Equations: Theory, ACM Trans. Math. Softw., 41 (2015), pp. 9:1-9:20.  PDF

DAESA Software

Algorithm 948: DAESA: A Matlab Tool for Structural Analysis of Differential-Algebraic Equations: Software, ACM Trans. Math. Softw., 41 (2015), pp. 12:1-12:14.  PDF 

Conversion Methods

Algorithmic Differentiation

How Algorithmic Differentiation Can Help Solve Differential-Algebraic Equations. Optimization Methods and Software, DOI: 10.1080/10556788.2018.1428605