About Me
Assistant Professor in Electrical Engineering,
Sharif University of Technology
Behzad Ahi was born in Tehran, Iran, in 1991. Since July 2023, as an assistant professor, he joined the Electrical Engineering Department of Sharif University of Technology, Tehran, Iran. Prior to joining the faculty, he was a postdoc researcher granted by Iran National Science Foundation (INSF) from July 2020 to March 2022. He is currently director of Data Fusion and Control Systems Lab.
Education
B.Sc
Electrical Engineering from the Sharif University of Technology, Tehran, Iran, (2009-2013)
M.Sc
Electrical Engineering from the Sharif University of Technology, Tehran, Iran, (2013-2015)
Ph.D
Electrical Engineering from the Sharif University of Technology, Tehran, Iran, (2015-2019)
Research Interests
Sensor fusion
Sensor fusion is the process of integrating data from multiple sensors to produce more accurate, reliable, and comprehensive information than what could be obtained from any single sensor alone. This technique leverages the strengths of different sensors, compensating for their individual limitations. For example, in autonomous vehicles, sensor fusion combines data from cameras, radar, and lidar to create a detailed and accurate representation of the vehicle's surroundings, enhancing safety and navigation. Deep learning and Kalman filters are the well-known tools in sensor fusion. By merging diverse data sources, sensor fusion improves decision-making in a wide range of applications, including robotics, navigation, tracking, healthcare, and environmental monitoring. In our lab, we focus on applications of sensor fusion in robotics and autonomous vehicles.
Autonomous vehicle
An autonomous vehicle (AV), is designed to navigate and operate without human intervention by utilizing a combination of sensors, cameras, radar, and artificial intelligence. Using a combination of sensors (e.g., radar, lidar, camera), AVs must accurately detect and respond to obstacles, pedestrians, and other vehicles, and handle complex scenarios (such as merging onto highways), ensuring safety in diverse and unpredictable driving conditions. In modular design viewpoint, the problem is break down into 5 category of navigation, perception, prediction, planning and control.
Reinforcement learning
Besides the supervised and unsupervised learnings, the reinforcement learning (RL) is a type of machine learning where an agent learns to make decisions by interacting with its environment and receiving feedback in the form of rewards or penalties. The agent aims to maximize its cumulative reward over time by exploring different actions and learning from the outcomes. This trial-and-error approach allows the agent to develop optimal strategies for complex tasks (e.g., playing hard games). However, we are interested in applications of RL in control of uncertain control systems.
Flight control systems
Guidance, navigation, and control (GNC) module is the main core in each flying vehicles (e.g, aircraft and drones). Navigation answers the question of "where the vehicles is in the word?". To answer this question, we need to integrate various sensors such as (accelerometer, gyroscope, GPS, magnetometer, vision, lidar, etc). Guidance answers the question of "How to get there?". To solve guidance problems, we use the well-known control techniques, such as those founded in adaptive and optimal control. The control strategy used in the inner loop of an aerial vehicles is called as "autopilot". In designing autopilots, we typically use the nonlinear control techniques.
Robotics
Robotics plays a critical role in advancing intelligent systems, enabling cutting-edge research that bridges control, perception, and real-world autonomous decision-making. Our laboratory is equipped with several advanced educational and research platforms, including the JetRacer AI, a high-performance AI-powered robotic car; a Robotis Engineering Kit 1 & 2 for rapid prototyping and RL algorithm development; and the Tello EDU quadrotors for aerial robotics research. We also maintain expert vision systems to support robotics applications in perception, sensor fusion, and real-time decision-making. We actively conduct research and development on a two-joint leg control system and the Unitree Go-2 quadruped robot located in a collaborating laboratory, as part of our robotics projects.
Courses
Digital Control Systems (Khatam university)
Linear Control Systems (25411)
Industrial Control Systems (25791)
Inertial Navigation (25442)
Optimal Control (25426)
Linear Control Systems Lab (25403)
Precision Instruments Lab (25404)
MS.c Students
Alireza Esmailnezhad
Mohammad Faraji
Mohammad Reza Gilak
Alireza Faghih
BS.c Students
Ali Mansouri
Amirali Pourdehghan
Mohammad Beiky
Taha Abedini
Erfan Bateni
Contacts
(+98)-21-66165982
ahi@sharif.edu
https://ee.sharif.edu/~ahi
3th Floor - 315 East, Department of Electrical Engineering, Sharif University of Technology, Azadi Ave., P.O. Box 1155-4363, Tehran, Iran