Ashkan Zehfroosh

About

Ashkan Zehfroosh is a multidisciplinary researcher whose work bridges robotics, formal methods, machine learning, and pediatric rehabilitation engineering. His research spans three interconnected domains: optimal motion planning for robotic systems, formal language-theoretic approaches to learning and inference, and intelligent systems for human-robot interaction in therapeutic contexts. Zehfroosh has made notable contributions to cable-suspended robot control, developing optimal path planning and feedback linearization techniques that account for dynamic obstacles and moving boundaries — work that has collectively garnered over 70 citations. His investigations into Signal Temporal Logic and Control Barrier Functions represent significant advances in reactive motion planning, offering computationally tractable alternatives to workspace discretization approaches that suffer from the curse of dimensionality. Perhaps most impactful is his work on the Grounded Early Adaptive Rehabilitation (GEAR) system, a smart environment designed for pediatric motor rehabilitation that has attracted 40 citations, reflecting strong interest from the clinical and engineering communities. Complementing this, his research on learning discrete interaction models from small data addresses a critical practical challenge in deploying human-robot interaction systems in rehabilitation settings. His 2018 paper connecting statistical relational learning with grammatical inference through model-theoretic string representations further demonstrates his remarkable range across formal and applied computational domains.

Research Focus

Key Achievements

Top Papers

1. 1
   Statistical Relational Learning With Unconventional String Models

   44 citations · 2018
2. 2
   Optimal motion planning of non-linear dynamic systems in the presence of obstacles and moving boundaries using SDRE: application on cable-suspended robot

   43 citations · 2014
3. 3
   GEARing smart environments for pediatric motor rehabilitation

   40 citations · 2020
4. 4
   Optimal path planning of a cable-suspended robot with moving boundary using optimal feedback linearization approach

   23 citations · 2014
5. 5
   Learning models of human-robot interaction from small data

   14 citations · 2017
6. 6
   Reactive motion planning for temporal logic tasks without workspace discretization

   10 citations · 2019
7. 7
   Non-Smooth Control Barrier Navigation Functions for STL Motion Planning

   9 citations · 2022
8. 8
   Optimal regulation of a cable robot in presence of obstacle using optimal adaptive feedback linearization approach

   6 citations · 2014
9. 9
   Learning option MDPs from small data

   5 citations · 2018
10. 10
    Control Barrier Navigation Functions for STL Motion Planning

    4 citations · 2022

Key Collaborators

Contact & Links