Arash Salemi

Arash Salemi

Robotics Engineer

Publications

Journal Articles

Toward personalized human-in-the-loop training: Real-time estimation of individual motor learning dynamics using the dual-rate model

Published in Computers in Biology and Medicine, 2025

This study presents a real-time estimation framework for modelling individual motor learning dynamics using system identification methods: Moving Horizon Estimation (MHE) and Extended Kalman Filter (EKF). The results show how these methods can enable adaptive and personalized robotic training, bridging control theory, neuroscience, and rehabilitation robotics.

Recommended citation: A. Salemi, A. Afkhami Ardekani, A. H. Vette, M. Nazarahari. (2025). Toward personalized human-in-the-loop training: Real-time estimation of individual motor learning dynamics using the dual-rate model. Computers in Biology and Medicine, 198, 111198. https://doi.org/10.1016/j.compbiomed.2025.111198
Download Paper

Work in Preparation

Optimizing Human-in-the-Loop Training: Real-Time Adaptive Task Scheduling via Model Predictive Control
A. Salemi, A. Afkhami, A. H. Vette, M. Nazarahari
In prepration, 2025.

This study introduces a control-theoretic framework that integrates Moving Horizon Estimation (MHE) and Model Predictive Control (MPC) to optimize task scheduling in human-in-the-loop robotic training. By leveraging real-time estimates of individual learning dynamics, the framework adaptively selects training tasks to maximize learning and long-term retention. The results demonstrate that MPC enables personalized and optimal scheduling, advancing real-time adaptive robotic training.