A Mechanical Engineering Graduate Student at Arizona State University. Graduated this May and currently looking for full-time opportunities. Skillsets include Robotics, ROS, Computer Vision, Python, C++ and Deep Learning.
-
Experience
Robotics Research Aide Jan’20-May’20
• Developed an accompanying middleware in ROS to operate a small autonomous vehicular robot in a physical testbed. Also, developed a GUI for the robot using PyQt5.
• Modelled the robot in SolidWorks and imported its URDF in Gazebo for simulation.
• Trained our YOLOv3 model for our custom dataset using TensorFlow to detect the objects present in the testbed with an accuracy of 97%.
-
Projects
Autonomous Surveillance Drone Jan’19 – Apr’19
• Used Robot Operating System (ROS) and Python for making the drone move autonomously and making it land in a specific spot by detecting Aruco markers using camera feed from the drone.
• Worked on You Only Look Once (YOLO) for object detection and sending only the required frames of interest to the master node.
Video link: https://drive.google.com/file/d/1UVYVGQczvnjHBOX2xTIamEPUnm7DHrdB/view?usp=sharing
Path Planning Optimisation of a mobile robot in a static environment Aug’19 – Dec’19
• Worked on Path planning optimisation of a mobile robot moving from an initial given point to a final point B with respect to static obstacles using Rapidly-exploring Random Tree – Star (RRT*) algorithm.
• Implemented Sequential Quadratic Programming (SQP) technique for determining the optimised path.
Trajectory planning KUKA Robot Sep’18-Dec’18
• Developed a mathematical model of 7-Degrees of Freedom (DOF) KUKA Robot using MATLAB. • Evaluated the Denavit–Hartenberg (D-H) parameters, performed forward and inverse kinematics using Jacobian Inverse method and performed the trajectory planning using Cubic polynomial and found the required joint pose and velocities.
• Validated the solution by simulating the experiment in MATLAB using Peter Corke’s Robotics toolbox and checked for obstacle collision.
Optimisation of a brake disc dimensions Sep’19 – Nov’19
• Performed Structural Analysis using ANSYS on the brake disc with respect to frictional and centrifugal forces to find its maximum von Mises stress.
• Performed DOE using Latin Hypercube Sampling and used Kriging method for response surface evaluation.
• Optimized the geometry using Non-Linear Programming by Quadratic Lagrangian (NLPQL) resulting in reduction of Equivalent stress (Von Mises) by 3.5% and Geometric volume by 20%.
Finite Element Analysis Projects Jan’19 – Apr’19
• Programmed a MATLAB code to analyse the performance of a 2-D steel fabricated cooling fin by determining its temperature and heat flux fields using linear, quadratic and cubic elements and computing the respective rate of convergence.
o Validated the MATLAB solution by computing the same problem using ABAQUS and comparing both the solutions.
o Used Richardson Extrapolation for conducting Mesh Refinement test by estimating the error for each analysed mesh.
• Performed FEA Analysis using ANSYS for a given Plexiglass design to find its ultimate tensile strength, factor of safety and drop height when subjected to a 10-pound mass.