Hello! I'm Soofiyan Atar

Researcher in the field of Robotics

I am a Graduate Student at the Unviersity of Washington Seattle pursuing Electrical Engineering Specilization in Robotics. I’m a research-oriented, team player, and looking forward to collaborating with an enthusiastic team or startup researching in the domain related to Robotics. I've participated in many robotics which includes Robocon Competition, Eyantra Robotics Competition, Eyantra Ideas Competition, and Texas Competition which are major national competitions, other than these I have also participated in various University level competitions and have also won almost every competition. My major projects are based on Swarm Robotics, Path planning, SLAM, Deep learning, Motion Planning, Control Systems, and Perception. I am currently working and researching algorithms related to swarm robotics. My goal is to continue researching in the field of robotics and learn more along the way.

Experiences

    IIT Bombay (Senior Project Technical Assistant)

    Worked on an autonomous X8 octo quadcopter with a payload of 15kg for spraying fertilizers and pesticides in agriculture. Dealt with Assistant robots for dynamic indoor environments using the custom mobile manipulator and developed the sim-to-real transfer framework. Managed projects for implementing a perception pipeline for harvesting tomatoes in agricultural environments and devising low computational algorithms for autonomous delivery robots. Designed two themes for eYantra Robotics Competition (eYRC) based on agriculture environments and assistant mobile manipulators.

    Supro Robotics LLC (Robotics Software Developer)

    Programmed a whole stack of autonomous fleet management for industrial robots used in effective warehousing and palletizing with a maximum carrying payload of 500kg. Built an autonomous navigation pipeline using the ROS2 framework with functionalities such as continuous progress checker, waypoint follower, and recovery behaviors. Employed dynamic obstacle avoidance for swarm robots using computer vision techniques such as collision risk assessment and selective decision-making.

Publications

These are my published research papers

    P2Ag: Perception Pipeline in Agriculture for Robotic Harvesting of Tomatoes

    Developed a perception pipeline in agriculture for harvesting tomatoes, mainly using mobile collaborative robots. Implemented a complete software stack compatible with Jetson TX2 and tested this pipeline on Gazebo. Technologies used: Semantic Mapping, Instance segmentation, Gazebo, ROS, MoveIt, and UR5 Robotic manipulator. Authored a paper on the same, which got published in the International Conference on Agents and Artificial Intelligence (ICAART) in November 2021, DOI: 10.5220/0010786100003116.

    LCPP: Low Computational Processing Pipeline for Delivery Robot

    Created low computational algorithms for autonomous delivery robots feasible on an embedded device and utilized efficient algorithms for segmenting roads, detecting obstacles, localization using HFM, and tracking objects. Technologies used: Efficient High-Fidelity Maps (EHFM), Semantic segmentation, Object detection, Object tracking, and Stereo camera (ZED2) Authored a paper on the same, which got published in the International Conference on Agents and Artificial Intelligence (ICAART) in November 2021, DOI: 10.5220/0010786300003116.

    Gaits Stability Analysis for a Pneumatic Quadruped Robot Using Reinforcement Learning

    Implemented Deep Q learning with pruned neural networks for input states to obtain a stable gait with minimal computation and higher efficiency. Authored and published a paper in the International Conference on Industrial Robots and Automation (ICIRA) in September-October 2021, Volume 15, Issue 09, ISSN: 1307-6892 (https://publications.waset.org/10012225/pdf).

    Amphibious Self-Balancing Autonomous Surveillance UGV

    Incorporated the entire software pipeline for balancing using LQR controller and integrating deep learning techniques for object detection and tracking, face recognition, threat assessment techniques, and chatbot. Authored and published a paper in the International Research Journal of Engineering and Technology (IRJET) in November 2021, Volume 08, Issue 10, ISSN: 2395-0072 (https://www.irjet.net/archives/V8/i10/IRJET-V8I10287.pdf).

    Learning Efficacy and Effect of Scaffolding in Online Engineering Education during COVID-19 Pandemic

    Explore the challenges of teaching "hardware-embedded Robotic Operating System (ROS) skills" online in a lockdown situation to participants through a hands-on competition. Studied the performance of teams based on several factors such as participant's score, task's difficulty level, number of resources given and the nature of online sessions by instructors. Authored and published a paper in the International Society of the Learning Sciences (ISLS), in June 2022, ISSN:1573-4552 (https://www.dropbox.com/s/ws5sdcfi72aykj1/ICLS2022%20Proceedings.pdf)

Competitions

These are all my Competitions which I have participated in, and I have contributed in all these organiazations.

    Robocon Competition

    This competition is an International Robotics Competition. I took part in this competition for 2 years. For the first iteration I was Firmware programmer. For the second iteration I was Senior Software programmer in which we implemented 4 legged Robot and also implemented different algorithms suitable for the particular theme. We stood 5th rank in this competition and also won Springer award.

    Eyantra Robotics Competition(eYRC)

    This Competition is a Robotics National Competition with 12000+ participants and we stood 3rd Rank for the Nutty squirrel theme. In this theme we had to apply Path planning algorithm with Line following algorithm and many other algorithms. We had to also design lift mechanism and arm for placing the nuts.

    Eyantra Ideas Competition(eYIC)

    This Competition is a National Competition where all the teams have to submit their ideas. We finished in top 21 teams in the final and our project was Autonomous Amphibious Surveillance UGV. In this competition we first got selected through Regional finals and then National finals with total 140+ teams competting at Regional finals.

    Texas Competition(IICDC)

    This was a training were we created a virtual lab in the field of computer security and also our lab was published on the Virtual labs official website. In this website we created a virtual lab using HTML, Javascript and CSS.

Internships

These are all my Internships which helped to boost my skills and also taught me concepts related to Robotics.

    Emobiliti systems(Robotics Software Engineer)

    Building Autonomous Industrial robots using ROS platform. Building custom algorithms as per the application. Autonomous guided vehicle for industrial use using stm32f7 and other sensors for feedback.

    Research Internship at IIT Bombay

    In this Internship, I implemented wireless programming for the ATmega2560 micro-controller using the ESP32 controller. We also implemented it for Arduino ide by adding custom boards for eYFi-Mega board and custom packages for the same. I also implemented wireless programming on Arduino ide without installing any other packages. Added custom board on GitHub thus can be used for any operating systems.

    Deep learning Internship at ATOS

    In this Internship, I implemented the deep learning algorithm to detect defects on wind turbines using drones. We used the Faster RCNN algorithm and also did data analysis on the data collected by us from various sources.

    Embedded Systems Internship at Eyantra

    This was an inhouse internship held by K.J. Somaiya College of engineering conducted by Eyantra. In this Internship, I implemented I2C communication protocols to interface the 6 axis gyroscope + Accelerometer sensor with AVR controller and also applied different filters(Complimentary filter, Kalman filter). Also implemented it on the stm32f4 microcontroller which is Cortex M4 based microcontroller.

    IoT internship at Faclon Labs

    In this Internship I implemented 3 major projects which were e-Riksha charging IoT based station which works on CAN for retrieveing battery data and then sending data to the server using MC60 microcontroller and using MQTT Protocol, other was to actuate a system using sms and also merging the data with the server at the same time, last one was to control the system using temperature sensor and sending the data to server.

Projects

These are all my projects which is done by me or I was a contiributor in these project:

Autonomous Mobile Manipulator for Tomato Harvesting

Implemented overall pipeline for harvesting tomatoes which includes Motion Planning, Perception, Robotic Manipulator and Localization. Remote Acess servers were also implemented to control the robot remotely. Implemented custom Tomato detection algorithm using Image segmentation technique on the Nvidia Jetson TX2 board. Further optimized algorithms to increase the effeciency as to work on real time data.

Autonomous Amphibious Self balancing UGV

Developed Self Balancing Amphibious Surveillance Robot which traverses autonomously and it is also made to be terrain proof. This bot has threat detection capability, audio and video live streaming, foot steps detection, object tracking, chat bot and dynamic path planning. This bot traverse using 2d Lidar, Depth camera, Camera, IMU and small Infrared sensors. This robot is controlled through Jetson nano and Atmega 328p which was integrated through ROS and also simulated using Gazebo and V-REP. DWA Planner and a combination of A* and Potential field path planning was implemented for optimal planning.

Autonomous Self Driven/Guided Vehical for Industrial purpose

This project was done in Emobiliti systems where the bot was autonomously computing path and also manual control was provided through android application. In this project major algorithms were ran through ROS and other supporting algorithms were written on C++ programming language. In this project hardware used was Lidar, IMU, and odometry sensors for autonomous navigation and mapping.

4 legged Autonomous quadrupled robot

Developed autonomous quadrupled robot which navigates using 2D Lidar, IMU and pneumatic actuator. For control we used LQR controller for each leg as to compute synchronized gait. We used STM32f4 micro controller for computation of all the algorithms. This was localized on a pre configured area which had known fences according to which we mapped the area in real time. As it was pneumatic based quadrupled robot it consumes very less energy.

Hacking Drones in the vicinity

Developed the drone and also algorithms to hack the drone which fly in the vicinity of any compound. We implemented PID controller for controlling the drone. To hack drones we used jammers and other approach was to bombared drones on the intruder drone. We also simulated it on Gazebo. We also designed LQR controller but was unable to implement.

Three wheel holonomic drive

This project was a part of Robocon Competition where we to collect the shuttlecocks and then drop it to the autonomous line follower robot. This drive uses omni wheels for traversing. This drive use to follow the fence and using odometry it use to localize itself and using IMU it use to orient itself. This robot was also control using ps2 controller for particular task. This robot was implemented on AVR controller.

Swerve Drive

This project was a part of Robocon Competition where we had to follow sinewave path using lasers. IMU and Incremental encoders which was done on stm32f4 microcontroller. This drive had to perform other task as well such as picking objects. It was controlled by ps2 controller for performing task, it use to traverse autonomously.

4 Wheeled robot Autonomous Navigation and Mapping

This project consist of a 4 wheeled robot with 2 drive wheels and 2 steering wheels. This project was done on ROS and Gazebo SImulator. The robot traverse autonomously in the environment. The bot uses Gmapping package for mapping and Navigation stack for autonomous navigation.

Honors and Awards

  1. 5th rank in ABU Robocon Competition(2018-19) held by IIT Delhi which is an International Competition but were not qualified for international round as we finsihed 5th place, also won springer award for finishing in top 8 teams.
  2. Winner in Eyantra Robotics Competition(eYRC 2018-19) held by IIT Bombay, where 12000+ students participated
  3. Finalist in Eyantra Ideas Compeition(eYIC 2019-2020) held by IIT Bombay which is a National level competition and finished in Top 21 teams for our Autonomous Amphibious Surveillance Robot
  4. Qualified for Semi finals in IICDC(India Innovation Challenge) by Texas Intruments which is a National level comeptition
  5. Winner in Find-a-way-fade-away competition which was held in K.J. Somaiya College of Engineering

Contact Me