Related Experiment Video
Updated: Apr 28, 2026

Real-Time Proxy-Control of Re-Parameterized Peripheral Signals using a Close-Loop Interface
Published on: May 8, 2021
Optimizing human-in-the-loop training: Real-time personalized task scheduling via model predictive control
Arash Salemi1, Amirhossein Afkhami Ardekani1, Albert H Vette2
1Department of Mechanical Engineering, University of Alberta, Donadeo Innovation Centre for Engineering, Edmonton, Alberta, T6G 1H9, Canada.
This study introduces a new real-time task scheduling framework using Moving Horizon Estimation (MHE) and Model Predictive Control (MPC) for motor learning. The new method significantly improves long-term retention and adaptation in training, outperforming traditional schedules.
Area of Science:
- Motor learning and control
- Robotics and human-in-the-loop systems
- Rehabilitation engineering
Background:
- Effective motor learning is crucial for rehabilitation and professional training, with task scheduling impacting multi-task learning.
- The dual-rate motor learning model highlights a slow-learning, slow-forgetting state crucial for long-term retention.
- Existing task scheduling methods lack optimality or real-time applicability due to rule-based or offline optimization approaches.
Purpose of the Study:
- To develop a novel, control-theory-based framework for real-time optimization of task schedules in human-in-the-loop training.
- To integrate Moving Horizon Estimation (MHE) and Model Predictive Control (MPC) for optimizing task sequences to maximize the slow-learning state.
- To enhance long-term retention and adaptation in motor skill acquisition.
Main Methods:
- Leveraged Moving Horizon Estimation (MHE) for real-time state and parameter estimation of the dual-rate motor learning model.
- Employed Model Predictive Control (MPC) to utilize MHE estimates for selecting optimal task sequences.
- Benchmarked MPC against an Alternating schedule and offline optimization in simulations and experiments (n=20).
Main Results:
- MPC demonstrated superior performance in simulations, achieving up to 27% lower cost function values compared to the Alternating schedule.
- Experimental results showed the MPC group had up to 46% lower mean motor errors in later testing phases.
- The MPC group exhibited significantly faster adaptation and stronger long-term retention (p < 0.05).
Conclusions:
- The developed MHE-MPC framework enables real-time, personalized optimization of training schedules for motor learning.
- This approach significantly enhances adaptation and long-term retention compared to traditional methods.
- The findings indicate a transformative potential for next-generation rehabilitation and high-performance skill acquisition.
Related Concept Videos
Open and closed-loop control systems
An open-loop control system operates without feedback from the output. It consists of two primary elements: the controller and the controlled process. The controller receives an input signal...
Reinforcement Schedules
Once a behavior is learned,...
PD Controller: Design
Designing a continuous-data controller requires selecting and linking components like adders and integrators, which are fundamental in Proportional,...
Control Systems
At the heart...
PID Controller