I pursue a PhD in the field of Machine Learning and Robotics with Jan Peters at the Robot Learning lab within the Empirical Inference dept by Bernhard Schölkopf. Before, I received a MSc degree in Biomedical Engineering at the Imperial College London working with Aldo Faisal. My BEng degree was rewarded to me in Information and Electrical Engineering from HAW Hamburg in conjunction with Siemens Healthcare.
My big aim is to use Machine Learning methods, especially a variety of different Reinforcement Learning approaches, to reach the performance of humans at fast changing, uncertain and high-dimensional tasks. For that purpose I develop algorithms that learn to control systems like the 7 DOF Barrett Wam arm or a self-designed muscle based robot to play table tennis.
For information and detailed construction details for the 4-DoF pneumatic artificial muscle actuated robot arm please send me an email: dbuechler[at]tue[dot]mpg[dot]de.
IEEE Robotics and Automation Letters, Robotics and Automation Letters, IEEE, 2018 (article)
Controlling musculoskeletal systems, especially robots actuated by pneumatic artificial muscles, is a challenging task due to nonlinearities, hysteresis effects, massive actuator de- lay and unobservable dependencies such as temperature. Despite such difficulties, muscular systems offer many beneficial prop- erties to achieve human-comparable performance in uncertain and fast-changing tasks. For example, muscles are backdrivable and provide variable stiffness while offering high forces to reach high accelerations. In addition, the embodied intelligence deriving from the compliance might reduce the control demands for specific tasks. In this paper, we address the problem of how to accurately control musculoskeletal robots. To address this issue, we propose to learn probabilistic forward dynamics models using Gaussian processes and, subsequently, to employ these models for control. However, Gaussian processes dynamics models cannot be set-up for our musculoskeletal robot as for traditional motor- driven robots because of unclear state composition etc. We hence empirically study and discuss in detail how to tune these approaches to complex musculoskeletal robots and their specific challenges. Moreover, we show that our model can be used to accurately control an antagonistic pair of pneumatic artificial muscles for a trajectory tracking task while considering only one- step-ahead predictions of the forward model and incorporating model uncertainty.
Our goal is to understand the principles of Perception, Action and Learning in autonomous systems that successfully interact with complex environments and to use this understanding to design future systems