The Delft Center for Systems and Control (DCSC) is looking for a PhD candidate to work on a 4-year project “Control and Generative AI for Human Digital Twin”. Controlling a digital twin of the human body with hundreds of muscles and joints to generate realistic human motion, e.g., walking, running, or jumping, is a challenging control problem. This is essential for conducting human-involved experiments in silico, significantly reducing the experimental cost in various applications, e.g., understanding human motor skills, designing rehabilitation programs, and designing assistive devices such as exoskeletons and prostheses.
Due to the complexity of human digital twin, this project will focus on developing novel control methods by exploiting reinforcement learning (RL) and generative AI. The developed approaches will be used for generating complex human motion across different populations, e.g., healthy individuals, individuals with mobility impairments, and individuals wearing assistive devices. While the primary focus of this project is on healthcare, the results will also have a profound impact on gaming, virtual reality, animation, and humanoid robotics. Experiments for collecting human motion data will be performed using a motion capture system in our Lab.
Deadline: May 11th, 2025
Application via the university website: TU Delft Website