Simulation environments for intelligent robotsDownload PDF

(3D Image Analysis and Synthesis)
Future robotic systems will operate in complex and constantly changing environments, such as cooperative production, private homes and offices. In this work, you will design a simulation environment for a robotic arm with a manipulator, as well as corresponding task controllers. The robot is specified as an actuated 3D model, i.e. configuration of the virtual robot is controlled by joint or motor positions. Manipulators (e.g. grippers) are adapted to the requirements of the task. The robot is placed into simulated environments derived from real-world problems, such as manipulation in home/office scenes, machine loading in production, packaging or sorting of small components.

You will work with the popular ROS framework and its state-of the-art simulation and visualization tools Rviz and Gazebo. This environment allows for the integration of many interesting algorithms, such as visual detection to find manipulated objects and the scene and physics simulation to consider gravity and inertia of objects. Detection and physics simulation are mainly addressed if this work is chosen as a master thesis.

This work is conducted within the RoVi project, which develops low-cost robotic systems with vision-based sensors. RoVi aims to transfer research results into products and a startup company.

Keywords: Robotics, ROS, 3D modelling
Requirements: ROS, C++/Python, Linux, 3D modelling