cv

• Outstanding Research Award, University of Pennsylvania
• 1st Prize in SICK LiDAR TiM$10K Challenge

• Gold Medal for securing 1st Rank, VJTI

Positioning and Controls Engineer

• Designing, Developing and Testing state estimation and control algorithms for a variety of heavy-duty vehicles.
• Performing root-cause analysis and collaborating with multi-disciplinary teams to provide swift solutions to on-site issues to reduce customer's downtime
• Developing and testing new algorithms and product ideas for the company to expand the product offering portfolio.
• Running the physical machines on the test sites to debug issues and test solutions.
• Common day-to-day tasks involve testing of algorithms and maintaining CI/CD pipelines.
• Skills
  • C++
  • Matlab
  • Simulink
  • Jira
  • Test Driven Development

Research Engineer

• Collaborating with a cross functional team to migrate the open source library for drones from ROS1 to ROS2.
• Skills
  • C++
  • Robot Operating System (ROS1 and ROS2)
  • Docker

Robotics Research Engineer

• Project 1 - Motion Planning and Control of swarm of mini-drones
  • Developed a simulator for a swarm of mini-drones which can fly in close proximity in an indoor environment.
  • Implemented a Hamiltonian path planning algorithm for the swarm of drones.
  • Currently performing physical drone tests to analyse the effect of down-wash (or wake) of the drones over one another.
• Project 2 - Motion planning of multi-agent micro-scaled robots
  • Developed a simulation for a swarm of micro-scaled robots, designing local-rules for a de-centralised multi-agent system using Python.
  • Developing and testing macro-scaled prototype of the micro-bot to understand the dynamics of the robot.

Data Scientist

• Developed Machine Learning models to perform customer segmentation, identify customer personas and predict future propensities for effective marketing.
• Improved the efficacy of the marketing campaigns resulting in a 20% increase in footfall for retail vendors.
• Developed PostgreSQL scripts for database management, data analysis and automating several monthly report processes for data visualization. Automation of these processes helped to save 40 hours of work per month.

Engineer

• Contributed to the development of CNC based programs and design and development of intricate mechanical products on machines like CNC, lathe, milling and grinding machines. Involved in designing of CAMs, indexers and complex mechanisms for automated production lines.

R&D Intern

• Designed and simulated a Solar Grid-tied inverter system using Matlab and Simulink.

Volunteeer

• Taught under-privileged kids Math and Science belonging to a wide-range of age groups.

MSE Robotics Engineering

• Relevant Coursework
  • Introduction to Robotics (Path planning and control of Robotic Arms)
  • Advanced Robotics (Motion Planning, State Estimation and Control of UAVs)
  • Learning in Robotics (State Estimation and Filtering techniques, Optimization and Reinforcement Learning)
  • Control and Optimization for Application in Robotics (Non-linear and modern control theories)
  • Feedback Control Theory
  • Performance Control and Stability of UAVs
  • Machine Perception
  • Machine Learning
  • Deep Learning
  • Mechatronics

Bachelor of Technology in Electronics Engineering

• Relevant Coursework
  • Electronic Circuit Analysis and Design
  • Microprocessors and Microcontrollers
  • Control Theory
  • Embedded Systems Design

SICK LiDAR TiM$10K Challenge

• Winners of the Robotics Competition organized by the company SICK.
• Designed and developed an autonomous solution for the public space sanitation industry, along with a team of 5 which is one of the 20 teams shortlisted in the USA.
• Skills involved- ROS, path planning, control, mechanical product design, mechatronics, electronic circuit design, perception.
• Final submission video link

Chance Constrained multi-agent Non-linear Model Predictive Control

• Formulated a non-linear model predictive control (NMPC) for a differential drive robot to perform decentralized path planning.
• Explored the trade-off of using hard constraints v/s barrier functions for smooth control transitions.
• Built the project in ROS 2, simulated in Gazebo environment, and used Drake as the mathematical framework to solve the optimization.
• Concepts explored - Model Predictive Control (MPC), Linear Quadratic Regulator (LQR) and Iterative LQR (iLQR)

Attention-based Networks for Human Trajectory Prediction

• Designed and trained a Transformer neural network (using PyTorch) to predict a human's trajectory on the TrajNet dataset.

Ensemble Kalman Filter

• Developed an Ensemble Kalman filter using a Neural Network (using PyTorch) which combined the filter outputs of an Error State Kalman filter, Complementary filter and Unscented Kalman Filter.

Autonomous VIO-based Quadrotor

• Developed and implemented the complete autonomy stack for a Crazyflie drone.
• Implemented A* path planning algorithm, Minimum-snap trajectory generation and a Geometric controller. These algorithms were tested on the actual drone as well.
• A Visual-Inertial Odometry (VIO) based state estimation and localization algorithm was implemented for the UAV as well and it was exhaustively tested in a simulation environment.

Simultaneous Localization and Mapping (SLAM)

• Implemented a mapping and localization algorithm with a particle filter using data from IMU and LiDAR sensor.

Franka Panda Robot Arm manipulation

• Used kinematics concepts to control and maneuver a Franka Panda robotic arm to perform tasks in a dynamic environment.
• Implemented a graph based path planning algorithm, RRT, to obtain a path for the 7 DOF robotic arm.

Semi-autonomous mobile robot

• Designed and developed a semi-autonomous mobile robot with multiple sensing systems to interact and perform complex tasks in a dynamic environment.