NIKHIL ARORA
716-***-**** ********@*******.*** linkedin.com/in/arora-nikhil www.nikhilarora05.com github.com/Nikhil9786 EDUCATION
University at Buffalo, The State University of New York Feb 2021 Masters of Science in Robotics Engineering
• Deep Learning • Introduction to Machine Learning • Robotics Algorithms • Computer Vision • Control Systems Thapar Institute of Engineering & Technology, India Jun 2019 Bachelors Engineering in Mechatronics Engineering
• Robotics Engineering • Artificial Intelligence • Automobile Engineering • Innovation and Entrepreneurship PROFESSIONAL EXPERIENCE
Software Engineer Intern, SimpL Labs INC., New Jersey, USA [Python, Swift, OpenCV] Jul 2020 – Aug 2020
• Developed computer vision model in python and swift using OpenPose for 15 exercises to detect human posture.
• Led a team of 3 to design a realistic 3D model of fitness center utilizing Google SketchUp and Enscape. Research and Development Intern, GS EngiTech, Ludhiana, India [SolidWorks, robot mechanics, actuators] Jan 2018 – July 2018
• Designed and programmed prototype of 6 DOF robotic arm using SolidWorks and Arduino.
• Achieved max payload of 100 grams and horizontal reach of 2 feet.
• Cut down cost of manufacturing of robot by 38.3% through careful material selection.
• Presented an idea to improve testing process of radiator pressure cap by minimizing time to 50% and increased safety. SKILLS
ML Stack: Python, Swift, C++, OpenCV, TensorFlow, NumPy, Scikit-learn, Pytorch, SQL, AWS Tools: Machine/Deep Learning Algorithms, Linux, Git Bash, GitHub, ROS, Jupyter, MATLAB, SAS, Linear Algebra, Probability Soft Skills: Interpersonal skills, problem solving, organizational skills, time management, leadership, people oriented PROJECTS
Self-Driving Car Simulation using Deep Learning [Python, DL, Nvidia model, Google Colab, Anaconda] Oct 2020 – Dec 2020
• Compared 4 neural networks and incorporated with NVIDIA model to train a car for autonomous driving on simulator.
• Decreased loss from 9% to 3% in training and tested car in new environment. Implementation of VGG16, ResNet18, InceptionV2 with CIFAR100 [Python, DL, Google Colab] Sep 2020 – Oct 2020
• Achieved higher training accuracies o with highest being 94.7% in ResNet18 and 62.3% testing accuracy in VGG16.
• Compared 3 with different regularizations and 2 optimizers, ADAM and SGD to report the best performance. Decision Making Machine Learning Model to Assist Judges [Python, Google Colab] Apr 2020 – May 2020
• Accomplished 2% higher accuracy and lowered societal cost roughly by $4,000,000 from existing COMPAS model.
• Compared 3 models to choose best from SVM, feed forward neural network, and Naïve Bayes classifier. Face Detection in the Wild [Python, OpenCV] Apr 2020 – May 2020
• Developed a computer vision model utilizing Viola-Jones Algorithm to detect faces from approximately 11000 images.
• Analysed 5700 faces of FDDB and 5000 non-faces images of CBCL dataset to train model. TurtleBot: Hand Gesture and Voice Controlled [ROS, Python, Linux, NLP] Oct 2019 – Dec 2019
• Incorporated ROS, NLP and Google APIs to get Turtlebot perform actions using hand gestures or voice commands. Climber Robot [SolidWorks, Arduino, Ultrasonic Sensors] Aug 2018 – May 2019
• Worked in a team of 3 to designed and program a robot able to climb up to elevation of 50 degrees.
• Accomplished to carry 200 grams of payload and achieved obstacle avoidance via real-time feedback from ultrasonic sensors. PUBLICATIONS
• Welding Using Different Robots- A Review – IJMPE Nikhil Arora, Lovepreet Singh & Navjot Singh Mangat DOI – IJMPE-IRAJ-DOI-15237 Mar 2019
• Design and Manufacturing Injection Mold Conformal Cooling Channels Using Additive Manufacturing - A Review – IJMPE Lovepreet Singh & Nikhil Arora DOI – IJMPE-IRAJ-DOI-15238 Mar 2019 VOLUNTEER EXPERIENCE
Volunteer Teacher, Har Hath Kalam, Patiala, India Aug 2016 – Mar 2017
• Displayed leadership qualities working as an English and Maths teacher to underprivileged kids of ages 6 to 10. Volunteer, Patiala Runners, Patiala, India Sep 2016 – Nov 2016
• Volunteered in team of 500 to organize marathon with approximately 4000 participants displaying teamwork abilities.