Erstellt am 16. August 2026
Student assistant(Hiwi) for Simulation and Reinforcement Learning for Swing-Suppression Control of a Lab-Scale Tower Crane
Technische Universität München
München, Bavaria 80333, Germany
Vollzeit
Reference: 569038521
Student assistant(Hiwi) for Simulation and Reinforcement Learning for Swing-Suppression Control of a Lab-Scale Tower Crane
14.08.2026, Student assistants, internships, student research projects
Tower cranes are underactuated systems in which rapid trolley, slewing, and hoisting motions can induce payload oscil-lations. These oscillations reduce positioning accuracy, increase settling time, and represent an important challenge for the development of autonomous and robotic crane systems.
Tower cranes are underactuated systems in which rapid trolley, slewing, and hoisting motions can induce payload oscil-lations. These oscillations reduce positioning accuracy, increase settling time, and represent an important challenge for the development of autonomous and robotic crane systems. In this project, we aim to develop a physics-based simulation and reinforcement-learning framework for swing-aware control of a lab-scale tower crane. The student will first further develop the existing crane simulation model in Isaac Sim. Particular attention will be given to the realistic representation of the suspended payload, pendulum motion, sensors integeration, actuator behavior, friction, and external disturbances such as wind. The physical lab-scale crane will subsequently be used for model identification and simulation calibration, enabling the estimation of important dynamic parameters and improving the fidelity of the simulation model. Based on the developed simulation environment, an Isaac Lab RL environment will be created to train a model-free controller capable of transporting the payload towarthe d desired target. Robustness against variations in payload mass, cable length, friction, wind disturbances, and modeling uncertainties will also be investigated through domain randomiza-tion. The final objective is to transfer the trained policy from simulation to the physical lab-scale tower crane.
Subject 1: Physics-Based Simulation and Model Identification
• Import and complete the tower crane URDF model in Isaac Sim.
• Model the crane actuation, payloaddynamics, and wind disturbances.
• Integrate sensors and update the simulation based on the real crane.
• Perform experiments on the lab-scale crane to identify key dynamic parameters.
• Tune and validate the simulation model against experimental measurements.
• Develop the simulation for further integrating the second robotis system
• Subject 2: Reinforcement Learning and Experimental Validation
• Develop the crane reinforcement-learning environment in Isaac Lab.
• Define observations, actions, reward functions, and training scenarios for accurate and swing-minimized pay-load transport.
• Train a model-free RLPlease send your CV, transcript, and, if available, information about previous experience with Isaac Sim/Lab, reinforcement learning, simulation, or robotics.transfer.
Timeline & Application Details:
Application Deadline: 15.09.2026
Active (in-person) participation is required.
Required Skills:
• Proficiency in RL, Isaac sim/Lab, C++/Python.
• Basic knowledge of dynamics and control theo-ry.
• Experience with real robots is a plus
Please send your CV, transcript and, if available, in-formation about previous experience with isaac sim/Lab, reinforcement learning, simulation, or robot-ics.
Kontakt: [email protected]
More Information
https://www.cee.ed.tum.de/ccbe/labs/robotic-fabrication-lab/
Crane Crane_topic_description, (Type: application/pdf, Size: 135.7 kB) Save attachment
14.08.2026, Student assistants, internships, student research projects
Tower cranes are underactuated systems in which rapid trolley, slewing, and hoisting motions can induce payload oscil-lations. These oscillations reduce positioning accuracy, increase settling time, and represent an important challenge for the development of autonomous and robotic crane systems.
Tower cranes are underactuated systems in which rapid trolley, slewing, and hoisting motions can induce payload oscil-lations. These oscillations reduce positioning accuracy, increase settling time, and represent an important challenge for the development of autonomous and robotic crane systems. In this project, we aim to develop a physics-based simulation and reinforcement-learning framework for swing-aware control of a lab-scale tower crane. The student will first further develop the existing crane simulation model in Isaac Sim. Particular attention will be given to the realistic representation of the suspended payload, pendulum motion, sensors integeration, actuator behavior, friction, and external disturbances such as wind. The physical lab-scale crane will subsequently be used for model identification and simulation calibration, enabling the estimation of important dynamic parameters and improving the fidelity of the simulation model. Based on the developed simulation environment, an Isaac Lab RL environment will be created to train a model-free controller capable of transporting the payload towarthe d desired target. Robustness against variations in payload mass, cable length, friction, wind disturbances, and modeling uncertainties will also be investigated through domain randomiza-tion. The final objective is to transfer the trained policy from simulation to the physical lab-scale tower crane.
Subject 1: Physics-Based Simulation and Model Identification
• Import and complete the tower crane URDF model in Isaac Sim.
• Model the crane actuation, payloaddynamics, and wind disturbances.
• Integrate sensors and update the simulation based on the real crane.
• Perform experiments on the lab-scale crane to identify key dynamic parameters.
• Tune and validate the simulation model against experimental measurements.
• Develop the simulation for further integrating the second robotis system
• Subject 2: Reinforcement Learning and Experimental Validation
• Develop the crane reinforcement-learning environment in Isaac Lab.
• Define observations, actions, reward functions, and training scenarios for accurate and swing-minimized pay-load transport.
• Train a model-free RLPlease send your CV, transcript, and, if available, information about previous experience with Isaac Sim/Lab, reinforcement learning, simulation, or robotics.transfer.
Timeline & Application Details:
Application Deadline: 15.09.2026
Active (in-person) participation is required.
Required Skills:
• Proficiency in RL, Isaac sim/Lab, C++/Python.
• Basic knowledge of dynamics and control theo-ry.
• Experience with real robots is a plus
Please send your CV, transcript and, if available, in-formation about previous experience with isaac sim/Lab, reinforcement learning, simulation, or robot-ics.
Kontakt: [email protected]
More Information
https://www.cee.ed.tum.de/ccbe/labs/robotic-fabrication-lab/
Crane Crane_topic_description, (Type: application/pdf, Size: 135.7 kB) Save attachment
