Working on developing Augmented Reality Technology for Dynamic Moving Systems.

Developed a grayscale modular vision-based pallet localization system for docking applications of autonomous mobile robots in indoor industrial settings that includes a 6 Degree of Freedom (DoF) pose estimation module and multi-object pose tracking module. Tested the developed proof-of-concept extensively for robustness and improved the system to address the shortcomings.

Worked on a 3D face error evaluation pipeline and created a Deep Learning pipeline for 3D face reconstruction in mobile phones

Lead the controls module (a team of four) towards developing a reliable, fast and performance-oriented software stack

Driverless | Software Core Team Member• Worked as path planning and high-level controls engineer• Developed an MPC controller for the Skidpad event to control the car at its lateral limits• Reviewed the previous seasons' works, identified the shortcomings and proposed improvements for the 2022 seasonAdditional Roles• Mechanical Repair team for Pilatus (car from the previous year)• Engineering Design Event Responsible• ForcesPro Solvers License Responsible

• Worked in collaboration with the Robotics Perception Group (under Prof. Dr. Davide Scaramuzza) and Sony• Worked on depth estimation using an event camera and a laser projector scanned at speeds up to 350 Hz, pushing the state-of-the-art.

Assisted with automated testing for CI/CD, writing unit tests, and code documentation for further development of battery optimization software under Prof. Dr. Christopher Onder at the Institute of Dynamic System and Control (IDSC)

• Trained and fine-tuned the then state-of-the-art object detection models for crop detection. Improved the performance of those models in edge-cases.• Developed a simulator visualizing the working of weeding blades, enabling efficient fault analysis of the robot's software pipeline.

• Worked as a part of the Surgical Robot Vision Group at the Wellcome / EPSRC Centre for Interventional and Surgical Sciences (WEISS) with Dr Agostino Stilli on Reinforcement Learning algorithm for a pick and place task in a medical environment.• The work was presented in Conference on New Technologies for Computer and Robot-Assisted Surgery (CRAS) 2020

• Developed multiple versions of Multi-Objective Bonobo Optimizer (MOBO), inspired from social behaviour of Bonobos, for solving multi-objective problems under Prof. Dilip Kumar Pratihar• Co-authored a paper based on my work which is published in the Journal of Knowledge and Information Systems