cv | Shamil Mamedov

June 2023 -
Content Writer on Deep Learning

deepschool.ru
- Writing articles on deep learning in Russian
Feb 2023 - May 2023
Visiting Researcher

RWTH Aachen University, Germany
- During my research stay in Sebastian Trimpe's group, I collaborated closely with Sebastian Trimpe and Rene Geist on the modeling of deformable linear objects, including ropes, cables, and long rods. Our work focused on the development of an interpretable model that merges concepts from the rigid finite element method with data-driven deep recurrent models.
2017 - 2021
Junior Robotics Researcher

Innopolis University, Center of Robotics
- Trajectory optimization and optimal control of pendubot and bipedal robot
- Dynamic model identification, external torque estimator design and collision detection for UR10E
- Modeling compliance of industrial robots under high loads

Physics informed machine learning
- Combining constituent physics model and function approximators for modeling complex physics such as flexible and deformable objects
Safe learning control
- Model-based controllers like MPC perform well provided good model but they are computationaly heavy for many application. Can we approximate them while providing safety guarantees?

Programming: Python, Matlab, C++
Robotics: Kinematics and Dynamics, Trajectory Optimization, Robot Operating System (ROS)
Machine Learning: JAX, PyTorch, Scikit-learn
Optimization: CasADi, Acados, IPOPT
Control: Optimal Control, Model Predictive Control, Nonlinear Control, Imitation Learning, Reinforcement Learning, State Estimation, Kalman Filter