Jobid=622440927249638286 (0.0988)
As part of a SNSF Co-Investigator Grant between ETH Zurich’s SoftRobotics Lab and University of Twente’s NeuBotics Lab, we are seeking a highly motivated postdoctoral fellow interested in advancing reinforcement learning approaches for the control of human-inspired musculoskeletal robots. The project focuses on creating digital twins of musculoskeletal robots equipped with neuronal control networks, with the aim of deriving robust robot controllers for sim2real applications.
If you are excited by interdisciplinary research at the interface of robotics, biomechanics, artificial intelligence, and neuroscience, we encourage you to apply.
The opening
The project combines advanced neuromusculoskeletal modeling, reinforcement learning, imitation learning, and robotic experimentation to enable the next generation of human-inspired musculoskeletal robots.
Your tasks will be
• Gradually adapt human neuromusculoskeletal models to incorporate robotic limbs based on muscle-like, variable-stiffness actuators.
• Development of digital musculoskeletal robot twins integrating neuromusculoskeletal models and models of electrofluidic robotic actuators.
• Develop digital twins of musculoskeletal robotic limbs equipped with muscle-like actuators and neural control networks.
• Use reinforcement learning (RL) to train digital robot twins to learn roboust movements underlying human-like joint impedance control.
• Develop imitation learning frameworks where robotic limbs learn to reproduce human-like movement and stiffness properties by observing a moving human twin.
• Sim2Real transfer on RL-policies to real hardware.
Your secondary tasks will include:
• Collaboration with interdisciplinary researchers in biomechanics, robotics, and machine learning.
• Dissemination of research through publications, open-source software, and international conferences.
About the Lab
The NeuBotics Lab is a multidisciplinary team at the forefront of neuromechanics, robotics, and human movement science. Our work bridges neuroscience, biomechanics, artificial intelligence, and robotics to develop adaptive control strategies and real-time biomechanical models for assistive and autonomous robotic systems.
You will join a dynamic research environment focused on translating computational neuromusculoskeletal models into real-world robotic applications, including prosthetic limbs, wearable exoskeletons, and autonomous musculoskeletal robots.
Why Join Us?
Your profile
Required Qualifications:
• A PhD degree in Robotics, Computer Science, Artificial Intelligence, Control Engineering, Mechanical Engineering, Biomedical Engineering, Electrical Engineering, or a related discipline.
• Strong publication record in robotics, neuromechanics, reinforcement learning, or related fields.
• Hands-on experience with:
• Excellent communication skills in English and ability to work in an interdisciplinary environment.
Knowledge of the following is a plus:
• Soft robotics or electrofluidic actuators.
• Prosthetic or assistive robotic technologies.
• Human movement analysis and biomechanics.
Our offer
We offer a position with a generous allowance:
Deel deze vacature: