Person

Emma Pabich

M. Sc.

Doctoral Researcher

Profile picture of Emma Pabich Copyright: © ,
Lehrstuhl für Data Science im Maschinenbau

Address

Building: 6050

Room: A 2.12

Dennewartstraße 27

52068 Aachen

Contact

WorkPhone
Phone: +49 241 80 92054
 

I did my Bachelors degree in Computational Engineering Science at RWTH Aachen University followed by a Masters degree in the same subject which I finished in May 2019. Afterward I worked as a research assistant at the Cybernetics Lab IMA&IfU of RWTH Aachen University. Since July 2020 I am working as a research assistant & PhD student at DSME supervised by Professor Dr. Sebastian Trimpe. I am working in the field of Reinforcement Learning, Learning-based Control and Robotics.

I am especially interested in how to train intelligent systems, e.g. robots, such that they master a variety of different skills. This will hopefully lead to robots that are intelligent in a similar way to human beings. Reinforcement learning is a way to teach robots their task autonomously. However, most reinforcement learning agents are trained on only one specific task which has to be predefined manually. In my research I want to develop methods to extract a task description automatically from images. My vision is to show a robot an image of his workspace and from this image the robot instinctively knows its task. This will enable to train a reinforcement learning agent able to generalize to a variety of different tasks.

Talks

„Industrial Reinforcement Learning: selbstlernende Prozesssteuerung für die Produktion“, VDMA Forum 4.0, Information Day on Intelligent Production Processes Research on Machine Learning and Artificial Intelligence, Frankfurt, Sep. 2019

Teaching

Teaching Assistant for the lectures “Informatik im Maschinenbau 2” and “Computational Science in Mechanical Engineering 2”

1 week seminar at the RWTH International Academy summer school “Robotics for Future Industrial Applications” (Jul. 2019)

2-day seminar organised by the RWTH International Academy on the topic of Model-Free Reinforcement Learning in Penang, Malaysia (Aug. 2019)

Supervised Student Projects

Jan Niklas Büscher, “Lung Cancer Segmentation with Guided Neural Networks for Computed Tomography Images”, RWTH Aachen University (May 2020)

Lukas Merten, “Learning Fair Trading in Financial Markets with Multi-Agent Reinforcement Learning”, RWTH Aachen University (ongoing)

Timo Bakr, Florian Berthhold, Timna Böttcher, “Deep Reinforcement Learning for Robot Navigation in Urban Environments”, RWTH Aachen University (ongoing)