Research Interests:

I am very attracted to robots with interesting and challenging dynamics: flying robots that perform challenging tasks, legged robots that can run, etc. In other words, my main interests include:

Aerial Robotics, Mobile Robotics, Robot Control, Trajectory Optimization, Underactuated Robotics, Legged Locomotion, Motion Planning.

 

Selected Research Experiences and Projects:

Note: This section will have a more comprehensive list of the projects by January 20th, 2017.

  • Research on "Autonomous UAV precision landing on a moving target" under supervision of Dr. Sebastian Scherer, Air Lab, Field Robotics Center, Robotics Institute, Carnegie Mellon University -
    February 2016 - Present
    More Information
    Currently developing a fast autonomous UAV that is able to locate, track and precisely land on a moving target attached to a ground vehicle without any communication with the ground vehicle or base station. As the student leader of the project, I am involved in all the parts of the project (control, estimation, perception, etc.)
     
    Indoor landing on a moving platform using visual servoing controlSimulation of MBZ International Robotics Challenge: Finding a moving vehicle, intercepting the vehicle and landing using visual servoing control
    Outdoor landing on a moving platform using visual servoing controlA new algorithm for real-time elliptical target detection
  • Development of a platform for "vision-based ground navigation research"; under supervision of Dr. Mansour Jamzad, Image Processing and Computer Vision Research Lab, Artificial Intelligence Group, Department of Computer Engineering, Sharif University of Technology -
    July 2012 - June 2014 (Intermittently)
    More Information
    I designed a low-cost car robot as a platform for vision-based outdoor navigation. The car was a low-cost toy car available at toy shops with maximum speed of 15km/s. It was equipped with a camera (replaceable with stereo camera for indoor navigation), odometer, speedometer, GPS sensor, ZigBee data transmitter/receiver, BlueBee bluetooth module, and a compass sensor. The car can be controlled using a base PC computer or an onboard android mobile, which receives sensor and video data from car and sends control commands to the car.
     
  • Implementation of "Extended Kalman Filter for State Estimation of a Loosely Coupled Visual-Inertial Aerial Vehicle ", Carnegie Mellon University -
    November 2016 - December 2016
    More Information
    This work is done as the final project of "16-831 Statistical Techniques in Robotics" course at Carnegie Mellon University.
     
  • Implementation of "Robot Localization using Particle Filter", Carnegie Mellon University -
    October 2016
    More Information
    This work is done as an assignment of "16-831 Statistical Techniques in Robotics" course at Carnegie Mellon University.
     
    Particle Filter Localization - Map 1Particle Filter Localization - Map 2
  • Research on "Human-robot interaction for a home robot" under supervision of Dr. Mansour Jamzad, Image Processing and Computer Vision Research Lab, Artificial Intelligence Group, Department of Computer Engineering, Sharif University of Technology -
    July 2012 - July 2014 (Intermittently)
    More Information
    Developed Human-Computer Interaction methods on a platform built by for Robocup @Home competitions. My tasks included face recognition, object recognition, mapping, localization and manipulation.
     

    The Service Robot platform (without the manipulator arm)
  • Derivation of "Minimum-time Trajectory and Control for a Quadrotor Intercepting a Ground Vehicle", Carnegie Mellon University -
    December 2016
    More Information
    In this project, a method for calculating the time-optimum (minimum-time) trajectory for a quadrotor to reach a ground vehicle moving with a known function is introduced. The method consists of two parts: an analytical part and iterative part. The analytical part finds a minimum-time trajectory for a quadrotor to reach a fixed final state from an initial state. The iterative method tries to find the solution of the whole problem by iteratively using the results from the analytic part. The quadrotor is assumed to have two controllers to control the yaw rate and acceleration.
    This work is done as the final project of "16-811 Math Fundamentals for Robotics" course at Carnegie Mellon University.
     
  • Research on "Persian and Arabic character recognition (OCR)" under supervision of Dr. Mohammad Eshghi. Two of my papers have been published on this work in ICEE'2009 and IEEE ICIT'2013 conferences -
    September 2008 - June 2014 (Intermittently)
    More Information
    The aim of this research was to implement a writer-independent OCR system for Persian and Arabic handwriting. I was the head of the research since September 2011. It also was the subject of my Electronics Engineering B.Sc. thesis.
     

    The proposed OCR system

    The set of primitive elements used in the proposed OCR system
  • Independent research on "Designing a fully automated traffic violation detection system using RFID lisence plates" -
    July 2014 - November 2014

    The proposed automated system for traffic violation detection
  • Independent research on "Power converters and inverters; using control techniques for Power Factor Correction". Two of my papers have been published on this work at IEEE PECON'2012 conference -
    March 2012 - September 2012

    Structure of the proposed inverter for variable inductive loads and low output power applications

    Schematic diagram of the SEPIC rectifier with the proposed PFC technique's circuit
  • Research on different heuristics for set-cover problem (B.Sc. Thesis) under supervision of Dr. Hamed Shah-Hosseini, Shahid Beheshti University -
    February 2011 - August 2011
  • Research Developer and Project Manager of "Unified SBU ACM Automation System and SBU Online Judge Project", Shahid Beheshti University -
    July 2010 - September 2011
    More Information
    The goal was to implement an automation system for most of regular needs in SBU ACM Student Chapter. I was Research Developer and Project Manager of the project in the specified period.
     

    Database design for the Unified SBU ACM Automation System and SBU Online Judge project
  • Research for implementing a "Combined Architecture for Stream Categorization and Intrusion Detection" for 2010 MEMOCODE co-design contest, Shahid Beheshti University -
    May 2010 - July 2010
    More Information
    The objective was to implement a high-speed regular expression matching system using Virtex-II FPGA. Although, we could not achieve the minimum required speed (500Mb/s) in time to participate in the 2010 MEMOCODE contest.
     

    Combined Architecture for Stream Categorization and Intrusion Detection
  • Implementation of "Spatial Pyramid Matching for Scene Classification", Carnegie Mellon University -
    Spring 2016
    More Information
    Implemented a scene classification system that uses the bag-of-words approach with its spatial pyramid extension.
    This work is done as an assignment of "16-720 Computer Vision" course at Carnegie Mellon University.
     

    Spatial Pyramid Matching for Scene Classification

Academic Memberships:

  • Institute of Electrical and Electronics Engineers (IEEE) -
    May 2012 - Present
  • IEEE Robotics and Automation Society (IEEE-RAS) -
    May 2012 - Present
  • Association for Computing Machinery (ACM) -
    April 2008 - April 2013
  • Chair of Shahid Beheshti University ACM Student Chapter -
    October 2011 - October 2012
December 23rd, 2016
- Air Lab, Field Robotics Center, Robotics Institute, Carnegie Mellon University