Prateek Sahay

•Created an autonomous rover model and enabled it in ROS (Robot Operating System) with multiple sensors (LiDAR and Camera) using their gazebo plugins including a differential drive controller plugin for actuation of the drive system. •Developed a C++based ROS package for this rover to track an object of a specific color by utilizing the information from the camera using a ROS service. •Performed Sensor Fusion of the rotatory encoders and IMU data on a Turtle-Bot by modifying the Robot Pose EKF (Extended Kalman Filter) package to estimate the 3D pose of the robot. •Localized the rover using the Adapted Monte-Carlo Localization (AMCL) package in ROS using particle filter technique to localize using a prior map of the environment. •Simulated Grid Based Fast SLAM on the Turtle-Bot using the gmapping ROS package. •Employed real-time loop closure using a GraphSLAM based approach known as Real-Time Appearance based Mapping by utilizing the RTAB-Map ROS package with a LiDAR and an RGB-D Camera to create a 2D and a 3D map of the environment. •Implemented path planning with the ROS navigation stack that uses a variant of Dijkstra's algorithm to plan the path. •Demonstrated the execution of the autonomous rover with a LiDAR and RGB-D camera to autonomously traverse an unexplored environment using SLAM and the ROS navigation stack. •Mapped a real-word workplace environment with RPLiDAR sensor using Hector SLAM package for ROS. Constructed an obstacle avoidance algorithm in python and performed real-world experiments using an RC vehicle. •Integrated Raspberry Pi, Sonar, servo controller and a keyboard with ROS to accomplish this task.

•Build a robotic work cell for automatic part placement of jet engine turbine blades using KUKA Iiwa LBR robot. •Simulated a pick and place operation for the KUKA iiwa robot using the ROS Moveit package to plan a collision free path in a cluttered environment. •Constructed a Python-based path planning algorithm for traversing the KUKA arm in a very cluttered environment model and simulated the same using ROS Gazebo and Rviz. •Performed real-world experimentation with the KUKA Iiwa robot and demonstrated collision free pick and place task

•Developed a robotic work cell with UR-10 at the center to inspect a device known as source holders for nuclear radiation leakages. •Programmed the robot to follow a planned path and perform a sequence of pre-planned events using the Urscript language. •Developed a python program to integrate the robot’s time stamp data with the nuclear radiation data coming from a Geiger counter attached to the robot using Modbus.

• Developed a program to demarcate lanes by drawing lane lines on video streams of a highway using OpenCV tools. • Developed a program to detect and track vehicles on a video stream of a busy highway using machine learning. • Developed a program to identify traffic signs in a picture using machine learning. • Trained a CNN to drive a vehicle autonomously on a highway in a simulation.

Specializing in the field of Robotics and Controls.

Topics: Memory Management: Smart Pointers, Multithreading and Advanced Object Oriented Programming