Pulkit Goyal Email & Phone Number

Sunnyvale, CA, US

Santa Clara, CA, US

Santa Clara, California, US

Menlo Park, California, US

Working on the hierarchy of motion planning for self-driving cars in an urban environment:Routing, Path Planning, Trajectory Optimization

• Researched and implemented Deep Learning model for object detection and segmentation on LiDAR data
• Integrated the trained network model in the ROS framework of the system

Pittsburgh, PA, US

• 𝗦𝘂𝗰𝗰𝗲𝘀𝘀𝗳𝘂𝗹𝗹𝘆 𝗱𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗱 𝗮 𝗽𝗿𝗼𝘁𝗼𝘁𝘆𝗽𝗲 𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗳𝗼𝗿 𝗮𝗻 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗱𝗶𝘀𝗮𝘀𝘁𝗲𝗿 𝗿𝗲𝘀𝗽𝗼𝗻𝘀𝗲 𝘀𝘆𝘀𝘁𝗲𝗺 𝘂𝘀𝗶𝗻𝗴 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗨𝗔𝗩𝘀 𝗮𝗻𝗱 𝗴𝗿𝗼𝘂𝗻𝗱 𝘃𝗲𝗵𝗶𝗰𝗹𝗲𝘀 𝘄𝗵𝗶𝗹𝗲 𝗮𝘃𝗼𝗶𝗱𝗶𝗻𝗴 𝗼𝗯𝘀𝘁𝗮𝗰𝗹𝗲𝘀.I worked on several aspects of the project:
• 𝗠𝗼𝘁𝗶𝗼𝗻 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴: Developed a lattice-based path and motion planning using ROS navigation for the AGV in the unknown environment
• 𝗕𝗲𝗵𝗮𝘃𝗶𝗼𝗿: Developed and implemented the behavioral model of the system using a Finite State Machine.
• 𝗟𝗼𝗰𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻: Implemented EKF to fuse odometry, GPS and IMU data for AGV localization
• 𝗢𝗯𝘀𝘁𝗮𝗰𝗹𝗲 𝗔𝘃𝗼𝗶𝗱𝗮𝗻𝗰𝗲: Contributed to setting up the ROS navigation stack for avoiding obstacles using a LiDAR

Roswell, Georgia, US

• Systems Design | System Specification | Architecture | Product Development
• Lead software developer for GreyOrange’s flagship product, Warehouse Sortation Systems, Recruited & setup team of 7 people, guided them in their individual research and coordinated & aligned efforts across varied.
• Worked closely with CTO, developers and managers throughout a product’s lifecycle – from its conceptualization to its multi-geography customer deployment.
• I initiated development of Peripheral Management System (PMS) for micro-controllers/sensors to change various system parameters using GUI which also included a warning generation mechanism based on location of each.
• I gained experience in the product development cycle by playing an important role in conceptualization, development, production & installation process setup, post deployment support and product improvisation throughout.

Roswell, Georgia, US

• Software Stack Design | Product Development | All Things Startup
• 𝗠𝗼𝘁𝗶𝗼𝗻 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴: Designed motion planning subsystem for differential wheeled robot(Butler) to navigate between grid points.𝗕𝗲𝗵𝗮𝘃𝗶𝗼𝗿𝗮𝗹 𝗺𝗼𝗱𝗲𝗹: Implemented behavioral model for task execution, safety system, and error.
• Developed a software layer in C & C++, to intelligently control actions/states of Sortation System and integrate embedded microcontroller modules with the customer APIs.
• Developed architecture to have an intelligent core module with a component-specific plugin for Sortation System.
• Developed a perception system for shipments with 1D & 2D barcodes, fed in high speed on the automated conveyor belts.

• 𝗪𝗼𝗿𝗸𝗲𝗱 𝗮𝘀 𝗽𝗮𝗿𝘁 𝗼𝗳 𝘁𝗲𝗮𝗺 𝗔𝗰𝗬𝘂𝘁, 𝗜𝗻𝗱𝗶𝗮'𝘀 𝗙𝗶𝗿𝘀𝘁 𝗜𝗻𝗱𝗶𝗴𝗲𝗻𝗼𝘂𝘀𝗹𝘆 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗱 𝗛𝘂𝗺𝗮𝗻𝗼𝗶𝗱 𝗥𝗼𝗯𝗼𝘁.Design of pathfinding, obstacle avoidance and foot-step placement algorithms for autonomous humanoids
• Researched and developed Path Generation algorithm taking into account the orientation of the robot after testing with Artificial Potential Mapping, A*, D* lite, Theta* using computer vision data and sensor feedback
• Developed pathfinding strategies for optimally navigating obstacle fields by exploiting the geometry of obstacles to minimize associated cost functions (decrease in robot speed / large variations in gait direction)
• Designed and implemented a soccer-playing behavior model of the robot using a Finite State Machine
• Using Computer Vision (OpenCV Lib) developed feature detectors and trackers for the soccer field and incorporated them into a Kalman Filter for accurate localization of Humanoid AcYut 4 for Autonomous Humanoid Soccer.
• Designed suitable kinematic structure for 28-DOF Humanoid Robot and used FEA to design high strength to weight ratio mechanical links for the robot head, torso, and limbs

IN

• 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗦𝗲𝗰𝘁𝗶𝗼𝗻, 𝗗𝗶𝘃𝗶𝘀𝗶𝗼𝗻 𝗼𝗳 𝗥𝗲𝗺𝗼𝘁𝗲 𝗛𝗮𝗻𝗱𝗹𝗶𝗻𝗴 𝗮𝗻𝗱 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 (𝗗𝗥𝗛𝗥)𝘋𝘦𝘷𝘦𝘭𝘰𝘱𝘦𝘥 𝘢𝘯 𝘰𝘣𝘴𝘵𝘢𝘤𝘭𝘦 𝘢𝘷𝘰𝘪𝘥𝘢𝘯𝘤𝘦 𝘢𝘯𝘥 𝘰𝘶𝘵𝘥𝘰𝘰𝘳 𝘯𝘢𝘷𝘪𝘨𝘢𝘵𝘪𝘰𝘯 𝘴𝘺𝘴𝘵𝘦𝘮 𝘧𝘰𝘳 𝘢 𝘜𝘎𝘝 (𝘜𝘯𝘮𝘢𝘯𝘯𝘦𝘥 𝘎𝘳𝘰𝘶𝘯𝘥 𝘝𝘦𝘩𝘪𝘤𝘭𝘦) 𝘸𝘪𝘵𝘩 𝘝𝘦𝘭𝘰𝘥𝘺𝘯𝘦 𝘏𝘋𝘓-32𝘌 𝘓𝘪𝘥𝘢𝘳.
• 𝗢𝗯𝘀𝘁𝗮𝗰𝗹𝗲 𝗔𝘃𝗼𝗶𝗱𝗮𝗻𝗰𝗲: Contributed in setting up the ROS navigation stack for obstacle avoidance using LiDAR
• 𝗩𝗶𝘀𝘂𝗮𝗹 𝗢𝗱𝗼𝗺𝗲𝘁𝗿𝘆: Implemented visual odometry model by applying ICP on features obtained from SIFT using R
• 𝗣𝗼𝘀𝗲 𝗲𝘀𝘁𝗶𝗺𝗮𝘁𝗶𝗼𝗻: Developed motion model using X-Sens IMU, Velodyne LiDAR’s IMU and GPS sensor along with the visual odometry model

Roswell, Georgia, US

• Designed & tested path correction and self-localization algorithms on automated guided vehicles (AGVs) utilizing feedback from camera sensors, optical encoders & ground markers for a robot-based warehousing solution.
• Implemented communication protocols (CAN, Ethernet, UART) for communication between motor drivers, drive control units, image processing unit, and master controller.

Master Of Science (M.S.), Robotic Systems Development [Mrsd], Mechatronics, Robotics, And Automation Engineering

Bachelor Of Engineering (B.E.), Mechanical Engineering

Senior Secondary Education (Class Xii), Science Stream (Pcm)

Secondary Education (Class X)