SUDARSHAN S HARITHAS
#** *** ***** *** ****** Jayanagar, Bengaluru, Karnataka 560011• Phone: +91-886-***-**** adlqxy@r.postjobfree.com https://sudarshan-s-harithas.github.io/index.html Education
Bachelor of Engineering (2016 – 2020): Electronics and Communication Engineering BMS College of Engineering - Bengaluru, Karnataka
CGPA: 9.19/10
Coursework: Computer Vision, Image Processing, Operating Systems, Advanced Microcontrollers (ARM7 and Cortex m3), Embedded Systems, Calculus, Linear Algebra, Discrete Mathematics, and Probability. Pre University (2014 – 2016)
Christ Junior College – Bengaluru Karnataka
Percentage- 94.50% (PCME)
SSLC: Class 10 ( 2014)
St Joseph’s Indian High School - Bengaluru Karnataka Percentage- 94.08%
Work History
[1] IIIT- Hyderabad, Robotics Research Center (Hyderabad, India) Research Intern: Sep 2020 – Present
I. Project1: Plane Based Visual Inertial Odometry estimation in Dynamic Environments Sep 2020- present
• Trained the PlaneRecover model for 3D planar reconstruction.
• Added a new inter-plane cost function to enhance the model performance.
• Implemented fully connected dense CRF (Conditional Random Fields) for improving segmentation results.
• Benchmarked performace of state-of-art VIO algorithms on various open source datasets. II. Project2: Autonomous Navigation of Aerial Vehicles Dec 2020- present
• Implementation of online mapping using Octomap and motion planning using OMPL (Open Motion Planning Library) for a quadrotor in hardware and simulation.
• Formulating approaches for mapping and planning considering environmental and sensor uncertainty.
• Assembly and offboard control of quadrotors using MAVROS and PX4.
[2] NCFlexe IIT Kanpur: TransPacks Technologies (Remote Internship) Computer Vision Intern: Sep– Dec 2020
I. Project1: CattleNet a system for cattle detection and identification.
• Obtained a stable accuracy of 98.0% using the ResNet50 CNN model for the FaceNet architecture.
• Used Adaptive Thresholding for Online classification of cattle.
• Trained YOLOv5 to detect cattle face, muzzle points, and eyes.
[3] Zen AeroLogiks - Bengaluru, India
Robotic Automation and Control System Intern: Feb– May 2020 I. Project1: Indoor Navigation System for mobile robots
• Developed a Generic Navigation system for autonomous navigation for unmanned aerial and ground robots, in an indoor environment using ROS
• Used Range Flow Technique to determine odometry information from Lidar scan data.
• Implemented Gmapping and AMCL for the mapping and localization process respectively.
• Implemented Path-Planning using Dijkstra’s Algorithm and DWA (Dynamic-Window Approach) as the global and local planners respectively.
[4] Dkiran Automation Engineering Services - Bengaluru, India research Intern: June - July 2018
I. Project1: Robotic AGV (Automated guided vehicle)
• Performed SLAM simulation for a Robotic AGV with Mecanum wheels, using the ROS navigation stack.
• Implemented Path planning for the AGV.
• Developed Python scripts for teleoperation of the robot during simulation through Bluetooth communication.
• Hardware construction and control of AGV using BeagleBone Blue. Projects
[1] ARDOP3.0 (HUMANOID ROBOT): July 2018 – July 2020 Project website: https://ardop3.github.io/
• Developed the manipulation and perception systems of ARDOP (Autonomous Robot Development Open Source Platform).
• Designed and constructed two 6DOF robotic arms from scratch.
• Created ROS simulation packages for ARDOP. Used Moveit with KDL (Kinematics and Dynamics Library) and OMPL (Open Motion Planning Library) to solve kinematics and motion planning.
• Additionally, solved inverse kinematics using the Jacobian pseudoinverse technique.
• Implemented using YOLOv3 and perspective camera transform for detection and localization of object in 3D space.
• Developed vision, logic, and system integration programs to enable ARDOP to play a game of Tic-Tac-Toe.
• Designed and constructed a custom power distribution circuit for the robot.
[2] GENERIK: May- Aug 2020
• GENERIK is a novel Inverse Kinematic (IK) solver.
• IK is solved as an objective function optimization problem.
• Conducted comparative studies choosing MSE (Mean Squared Error) and LogCosh (Logarithm of the Hyperbolic Cosine) as objective functions.
• The solver compatible with a variety of robotic manipulators and is capable of providing solutions with accuracy in the millimeter range
[3] SERP (Swarm Embedded Robotic Platform): Feb – Apr 2019
• SERP is a low-cost miniature-sized differential drive robot designed for swarm robotic research.
• Developed modular middleware libraries (C/C++) to support Arduino SDK for STM32.
• SERP features multiple sensors and has a USB interface for PC connectivity, Bluetooth, and Wifi modules for wireless communication.
[4] ARDOP2.0 (HUMANOID ROBOT): Mar-May 2018
• Processed IMU and wheel encoder data to obtain odometry information.
• Developed Python scripts for speech to text and text to speech conversion using GTTS
[5] Planter Bot (E-yantra National Level Robotics competition 2017): Aug 2017 - Feb 2018
• Designed and constructed a differential drive robot, with sensors such as quadrature encoders and Pi-cam interfaced with Raspberry pi.
• Developed algorithms for the robot to detect and trace white or black lines using OpenCV
• Performed Image Processing operations such as shape detection, edge detection, contour detection, overlay of multiple images (images of flowers) on a single host image(image of the farmland).
[6] Mini Projects
• Forgery Detection In Handwritten Signatures
• YouTubeSpam Comment Detector
• Balance Bot (2 wheel robot)
Publications
[1] “RP-VIO: Robust Plane-based Visual-Inertial Odometry for Dynamic Environments”, International Conference on Intelligent Robots and Systems, IROS 2021 (under review)
[2] “ARDOP: A Versatile Humanoid Robotic Research Platform”, Advances in Robotics, AIR 2021 (under review)
[3] “GENERIK: An Objective Function Optimization Approach to Inverse Kinematics”, 6th IEEE International Symposium on Smart Electronic Systems (IEEE- iSES), 14-16 December 2020.
[4] “GenNav: A Generic Indoor Navigation System for Mobile Robots”: 4th International Conference on I- SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC) (IEEE I-SMAC). 7-9 October 2020.
[5] "Introduction to Gravitational Waves and Detection Principles", a dissertation submitted to M.P Birla Institute of Fundamental Research.
Courses
[1] Fundamentals of Reinforcement Learning: Course by University of Alberta on Coursera
[2] Self-Driving Cars Specialization A set of 4 courses by the University of Toronto on Coursera
[3] Deep Learning Specialization: A set of 5 courses by deeplearning.ai on Coursera
[4] Machine Learning with Python-From Linear Models to Deep Learning: (audited) Course by MIT on edX
[5] 100 hours Certificate Course in Astronomy and Astrophysics: M.P Birla Institute of Fundamental Research. Skills
• Programming Languages: Proficient: Python, C Intermediate: C++ Basics: Matlab, Latex, HTML
• Software: Robot Operating System (ROS), MAVROS, Tensorflow, PyTorch, Keras, PX4, ArduPilot Operating Systems: Linux, Windows Gaming Engines: Unity3D, Unreal.
• Hardware: Nvidia Jetson, Raspberry Pi, BeagleBone Blue, STM32 BluePill, Arduino (All boards), NodeMCU ESP8266, Wi-Fi and Bluetooth chipsets, XBee, Intel 8051, familiar with Laser Cutting, 3D printing, and usage of hand tools.
Co-curricular Activities
• BMSCE Startup week 2018 – Procured Seed money to bootstrap our intended start-up called Polaris Labs at the BMSCE start-up week. At Polaris Labs, we develop various technological solutions to autonomous robots.
• Fury Road V2 (Phase Shift 2016): Secured first place in an RC car (Bluetooth Controlled) construction and racing competition.
Workshops/Events organized:
• A member of the organizing committee of App Controlled Car using Arduino, more than 180 students were trained on the course three workshops.
• A member of the organizing committee of Fury RoadV3 (RC car racing competition) the flagship event of Phase Shift (2017).
References:
Harish V Mekali Dr. Geetishree Mishra
Assistant Professor Assistant Professor
Dept of ECE Dept of ECE
BMS College of Engineering BMS College of Engineering Bengaluru 560 019 Bengaluru 560 019
Email: adlqxy@r.postjobfree.com Email: adlqxy@r.postjobfree.com