Personal and Educetional
Projects

My main task as a Robotics and Machine Learning researcher at Sber Robotics Lab was developing real-time algorithms for safe indoor robots' navigation.
One project was to develop neural networks that are capable of detecting glass and Mirrors in real-time (Which are very hard to detect using traditional sensors)

• More details about the project

My master's degree project was "Collective Exploring Using a Swarm of Quad-copters". The main goal of the project was to achieve Collaborative Safe Exploration of new environments using a swarm of quadcopters. During working on this project I worked on the following subtasks
- Developed a collision avoidance algorithm and published a paper about this algorithm.
- Developed an improvement for the path planning RRT algorithm for improved performance in large complicated environments. I also published a paper about this algorithm (More details on the Research page)
- Applied a Machine Learning model to control the quadcopters. - Developed an algorithm for dividing the exploration space between the quadcopters.
- Applied SLAM algorithm for all quadcopters.
- Developed an algorithm for merging maps from different quadcopters to get the final map.
Note: This project was planned to be applied and tested on real quadcopters, but due to the COVID-19 pandemic it was only performed and tested in a simulator.

My fifth-year project was building a quadcopter, in which I fully designed the frame of the quadcopter and 3D printed it. The main controller of the quadcopter was a beaglebone black controller.
In addition, I built a 3-DoF testing platform for testing the controllers.

During my bachelor's degree at HIAST university ( Higher Institute for Applied Sciences and Technology) I have completed two academic projects, the first of which is my fourth-year project "3-Finger Adaptive Robot Gripper". In this project, I fully designed the gripper and manufactured the parts in the university's workshop. I also designed the electrical circuit that controls the gripper. Using C# I created a user interface that includes all functionalities of the gripper such as (controlling individual motors (position, speed), controlling the shape of the gripper (freestyle, or choose from multiple predefined shapes), etc.) I added visual sensing using a camera which was able of estimating the dimensions of objects (and decide the best grip to catch each object)
I also built a glove for imitating the movement of a human hand.

The project's video on YouTube.

During my bachelor's degree, I participated in four WRO competitions (Advanced category). In all four, my team was ranked first in the national competitions and was qualified to participate in the world finals.

This project was the start point of me working on AI projects.
In this project, I built a simulator for showing the natural selection process in the game of snake, where the goal of the game is to get the maximum number of food objects without crashing.
The main idea is to start with a generation of snakes with random neural networks (minds) which decide how each snake behaves.
After each generation, snakes of the next generation are a mutation of the best snakes of the last generation. In the simulator, it is possible to see the activity of the neural network of each snake and see detailed information about it, such as at which generation it was created, who is the father of this snake, and other information.
More details about this project on my Github repository

• Github Repository

Chess Project

A program for detecting the chessboard in an image and detect the positions of the pieces

Telegram bot for changing image style

I Programmed a telegram bot for applying style transfer on user images.
The user can choose the style image and the image to change.

• Github Repository