GET-Forschungsseminar Abstracts
Master's Thesis Presentation: Autonomous Robot Navigation in Real Environments Using Reinforcement Learning
Asjad Sohail, GET Lab
Presentation: 17.03.2021, 16:30h
Abstract:
Autonomous robots are getting increasingly popular these days because they are used more and more for everyday tasks, such as cleaning floors or transporting goods from one place to another. The ability to navigate in a cluttered environment while avoiding obstacles is a fundamental ability that a robot must possess. In addition to many existing traditional methods, such a behavior can also be learned with the help of reinforcement learning. In this thesis robot navigation based on LIDAR data is realized through a deep reinforcement learning method. First, different models and reward functions have been used to train the robot in virtual environments. Finally, the modle showing the most promising results has been tested in the real world.