Hamidreza Raei

<about>

Research Interests

Reinforcement Learning

Developing RL algorithms for sim-to-real transfer, non-prehensile manipulation, and locomotion control on quadruped robots.

Computer Vision & ML

Multi-modal sensor fusion (LiDAR, Camera, IMU), visual odometry, and deep learning for autonomous systems.

Data-Driven Control

Model-based and model-free control architectures for mobile collaborative robots and UAVs.

Academic Background

PhD in Electronic, Information and Bio-engineering

Polytechnic University of Milan

Sep 2023 – Present GPA: 29.66/30 Marie Skłodowska-Curie Scholar

Focus: Reinforcement Learning for Robotics & Human-Robot Interaction

MSc in Mechanical Engineering

Gwangju Institute of Science and Technology

Aug 2020 – Aug 2022 GPA: 3.63/4.0 Full Scholarship

Thesis: Camera and LiDAR Fusion for Autonomous Landing on Moving Targets

BSc in Mechanical Engineering

Isfahan University of Technology

June 2020 Last 2yr GPA: 3.46/4.0

Thesis: Controller Design for Ball and Beam System using Time-Response Analysis

<publications_projects>

Recent Research & Projects

Submitted AIM 2026

Sim-to-Real RL for Ball-Balancing Locomotion on Quadruped Robots

Developed reinforcement learning framework enabling quadruped robots to perform dynamic ball-balancing locomotion with zero-shot sim-to-real transfer using domain randomization.

Reinforcement Learning Sim-to-Real Quadruped MuJoCo

Accepted RA-L 2025

RL Approach to Non-Prehensile Manipulation Through Sliding

Actor-critic DDPG framework for sliding manipulation with LSTM-based real-time friction estimation. Achieved zero-shot sim-to-real transfer on hardware platforms.

DDPG LSTM Non-prehensile Zero-shot Transfer

BioRob 2024

Multipurpose Interface for Close/Far-Proximity MCR Control

Designed visuo-haptic interface with detachable mechanisms and visual-inertial odometry for mobile collaborative robot control, matching commercial MoCap performance at lower cost.

HRI Visual Odometry MCR ROS2

ASCC 2022

Autonomous Landing on Moving Targets via Sensor Fusion

Developed 30Hz relative localization system using 1D LiDAR, camera, and IMU fusion. Implemented PID, fuzzy-PID, and LQR controllers for UAV trajectory tracking and landing.

LiDAR Camera IMU Kalman Filter PX4

Project

Speed Estimation using Dash Camera & CNN

Leveraged comma.ai dataset and OpenCV optical flow with TensorFlow CNN for vehicle speed estimation. Achieved validation MSE ~7.56 with custom architecture and dropout tuning.

TensorFlow OpenCV Optical Flow CNN

Project

Ball & Beam Control & Fuzzy-PID Auto Tuner

Engineered precision ball positioning system with servo motor, potentiometer feedback, and Arduino PID control. Developed MATLAB fuzzy logic auto-tuner for PID gains.

MATLAB Simulink Fuzzy Logic Arduino

<demos>

Research Demos & Videos

No video loaded

📁 Place your video file in: ./videos/
📝 Update the <video src=""> tag in the HTML
🎬 Format: MP4 / WebM | Max recommended: 50MB

Work Experience

Dec 2022 – Present

Research Fellow

Human-Robot Interfaces and Interaction Lab (HRI²) @ IIT

Developing RL-based manipulation and human-robot collaboration interfaces. Leading projects on sim-to-real transfer and visuo-haptic teleoperation.

Sep 2020 – Nov 2022

Research Assistant

Dynamics and Control Lab @ GIST

Worked on UAV localization, multi-sensor fusion (LiDAR/Camera/IMU), and advanced control systems (PID, LQR, Fuzzy) for autonomous landing on moving platforms.

Jan 2018 – Aug 2018

Engineering Intern

Shahid Mohammad Montazeri Power Plant

Gained practical experience in power plant operations, control systems monitoring, and industrial automation workflows.