Post-doc: Predictive Control of Leg Muscle Dynamics via Wearable Exoskeletons

Post-doc: Predictive Control of Leg Muscle Dynamics via Wearable Exoskeletons

University of Twente

Enschede, Netherlands

Are you passionate about creating next-generation wearable exoskeletons that can actively regulate muscle-tendon loading and prevent musculoskeletal injury risk? Are you interested in developing predictive, closed-loop control strategies that combine artificial intelligence, real-time musculoskeletal modelling, and model predictive control, operating at the time scale of human neuromuscular dynamics (milliseconds)?

We are seeking for a highly motivated postdoctoral researcher to develop real-time, predictive control frameworks for ankle exoskeletons that regulate calf muscle-tendon forces during human locomotion. A central goal is the short-term (millisecond-scale) estimation and prediction of muscle activation and tendon force over a prediction horizon (e.g., 200 ms), with validation using ultrasound measurements of the Achilles tendon. The position is at the intersection of musculoskeletal biomechanics, AI-based prediction, and real-time exoskeleton control.

Research Focus

The successful candidate will work on the development of predictive, real-time control frameworks for ankle exoskeletons that regulate muscle–tendon forces during locomotion.

This includes:

• Developing and calibrating real-time musculoskeletal ankle models (using CEINMS-RT), with emphasis on the Achilles tendon.
• Combining AI-based prediction (e.g., TCNN, LSTM, etc) with musculoskeletal models to estimate and predict muscle activation and tendon force over short horizons (e.g. ~200 ms).
• Integrating these predictions into model-predictive control (MPC) or reinforcement learning (RL) frameworks to compute optimal exoskeleton assistance in real time.
• Validating the developed methods in human experiments using motion capture, electromyography, ultrasound, and dynamometry.

Your Tasks

• Developing and validating subject-specific musculoskeletal ankle models.
• Combining AI-based prediction with MSK models for real-time muscle-tendon force estimation.
• Designing predictive control strategies that regulate muscle-tendon loading via wearable exoskeletons.
• Implementing and testing control algorithms in simulation and real-time settings.

Your profile

Required Qualifications:

• PhD in Robotics, Control, Mechanical Engineering, Computer Science, or a related discipline.
• Experience with:
  • Model predictive control and/or reinforcement learning
  • Musculoskeletal or biomechanical modelling
  • Control of wearable robots or exoskeletons
  • Real-time programming (C++, Python)
• Knowledge of the following is a plus:
  • Real-time communication systems (e.g. EtherCAT, CAN bus)
  • Closed-loop control of robotic systems
  • Experience with experimental human movement data (EMG, ultrasound, motion capture)

You are creative, proactive, and comfortable working at the interface of AI, physics-based modelling, and control.