Wojciech Turlej

janowy2

Automatics and Robotics student (AGH, Poland), robotics researcher, programmer, electronics hobbyist. Research interests span the areas of humanoid robots, optimal walking, trajectory generation, artificial intelligence in robotics.

In free time craftsman, drawer/painter, mountain hiker, guitar player.

Feel free to ask me any questions – you can contact me by e-mail or linkedin.


Robotics projects:

flips_anim

Novel trajectory generation method for compliant bipedal robots. The method allows to generate and optimize a motion that fulfills given task with minimum heuristics or manual adjustments. Resulting trajectories utilize elastic structure of the robot to perform very dynamic, energy-effective motions.

On the animation – example of produced motion for backflip task. Initial small jump was found by the method autonomously and allows to inject additional energy to the system in order to perform sufficient jump.

Project carried out in the German Aerospace Center (DLR) under Alexander Werner’s supervision. Paper submitted for IROS 2017.


Neuroevolutionary control of bipedal robots

neuroevo_anim

Novel approach for control of mechanical systems, in particular bionic robots, including bipedal robots. Presented methodology simplifies process of creating and optimizing control systems – controller based on neural networks and finite state machines is automatically optimized by evolutionary algorithms.

Described approach allows to obtain dynamically stable gait cycles on underactuated robots. Resulting controller actively stabilizes the system in presence of external disturbances. Optimal controllers for bipedal robots of various sizes and masses are derived without any external guidance.

Bachelor thesis, awarded as Faculty’s Best Thesis, presented at 52. Scientific Circles Conference in Cracow.


lib_anim2

Work with Alexander Werner and Gianluca Garofalo in the German Aerospace Center.

Efficient C++ library for robotics computation, simulation, visualization. Used, among others, for Toro humanoid robot and C-Runner compliant bipedal robot. Implements recursive constrained inverse/forward/hybrid dynamics algorithms, collision detection, realistic Featherstone’s contact model, 3D visualization, URDF parser, Matlab and Python wrappers and many other features. Thanks to templated structure and the automatic differentiation, the library can be used for fast computation of all components of the dynamics equation with their derivatives.