The Experiential Learning Ventures
ROBOTICS
Experiences
Robotics Software Engineer
Vecros Technologies Ltd.
New Delhi, India
June '23 - April '24
• Engineered scalable camera and LiDAR-based SLAM systems in a 3-member team, enabling real-time mapping for mines and warehouses
• Designed and implemented an end-to-end motion planning pipeline with dynamic obstacle avoidance, deploying it in both simulation and on drones
• Leveraged expertise in system design, operating systems, computer vision, parallel computing (CPU/GPU),
optimization algorithms, estimation filters, and software engineering to ensure performance and reliability
• The drone solutions were featured on Shark Tank India, Season 3 which had a viewership of over 12.5M
Lidar based realtime 3D mapping system The experiment demonstrates the Lidar based SLAM and localisation capabilities of the drone in a GPS denied environment. The drone solely relies on the Lidar and the algorithms onboard to stabilise and move around which creating a realtime map of the environment
Camera based autonomous localisation system The experiment demonstrates the camera based visual inertial odometry capabilities of the drone in a GPS denied environment. The drone solely relies on the camera and the algorithms onboard to stabilise and move around.
Research Intern
Multi-Robot Systems Group, CTU
Prague, Czech Republic
August '22 - December '22
I was fortunate to be invited to the MRS group at CTU in Prague for my undergrad thesis. It was my first trip to Europe where I -
• Developed a Deep Reinforcement Learning motion planning algorithm for drones with a 3-member team
• Implemented a Deep Q Network on a custom dataset generated using geometrical path planning methods
• Extended the lab’s ROS framework to enable real-time drone control using PyTorch models
Please checkout the github repos for the implementation deep dive - draca_ros || draca_planner
(draca - Deep Reinforcement learning based Collision Avoidance)
Projects
Image Segmentation using Attention & State Space Models
Deep Learning Course Project, WPI
Worcester, USA
October '24 - Present
• Part of a 4 member team to develop a state of the art organ semantic segmentation model for CT scans
• Integrated Super Token Attention with SSM-based Mamba architecture into a U-Net framework, achieving a DICE score of 87.02% on the multi-organ Synapse dataset
Ensemble Learning for robotic grasping
• Developed novel ensemble networks to integrate pre-trained models from diverse domains, achieving superior performance across all domains compared to individual models
• Achieved current state of the art performance in robotic grasping ensemble networks by improving mean
IoU by 7% over the best performing individual model. MER Lab grants funding to continue this research
MER Lab, WPI
Worcester, USA
August '24 - Present
SwarmMap - Wifi map sharing and simulator for swarms
Inspire Lab, BITS Pilani
Pilani, India
January '22 - May '22
I developed a cutting-edge WiFi-based communication package for mobile robots with the guidance of a Professor, as an independent research project. Using Java libraries, I created a GUI to spawn robots and obstacles on click, for testing. The robots could share a global map with each other in real-time. I tested the platform and the algorithm by moving the robots using A*.
Comparison of Localisation algorithms
Autonomous Mobile Robotics Course Project, BITS Pilani
Pilani, India
November '21 - December '21
Our 3 member team undertook a course project that compared two cutting-edge localization techniques - the Advanced Monte Carlo localization technique and Dead Reckoning - by leveraging the IMU and wheel odometry using EKF. As one of the team members, I had the pleasure of delving deep into the intricacies of Kalman Filters and the Monte Carlo method, while implementing the simulations in ROS, Gazebo & Rviz to rigorously test the algorithms.
Technical Teams
The Space Crew
Localisation Head
Worcester, USA
October '24 - Present
Space robotics is a fascinating field and I am currently working with fellow space enthusiasts to develop autonomy solutions for the Lunar Surface.
​
I am currently leading the localisation efforts wherein we will use monocular depth estimation and visual inertial odometry to provide position feedback to our Lunar rover
​
This is part of the Lunar Autonomy Challenge by NASA
​
Consortium for Research in Space Systems (CRISS) Robotics
Arm Programming Lead & Core member
Pilani, India
March '21 - May '22
Inspired by the landing of the Perseverance rover by NASA, all the space enthusiasts at my college gathered to make prototype Mars rovers to participate the University Rover Challenge. I was part of the robotic arm team where I -
• Led the development of a 5 DoF robotic arm for a Mars Rover using ROS, Gazebo, RViz and MoveIT, enabling various manipulation tasks like flipping switches, pulling ropes, tightening screws, etc.
• Mathematically modelled the inverse kinematics and dynamics of the arm in MATLAB
• The team won the International Rover Design Challenge 2023 in its maiden attempt
MISCELLANEOUS
Summer Intern
Bajaj Auto Ltd.
Remote
June '21 - July '21
As a sophomore, I interned at Bajaj Auto, a renowned automobile company in India. My project involved analyzing pressure data for two IC engine models using Python and related libraries to determine heat release and thermal efficiency at different RPMs. This immersive experience deepened my appreciation for IC engine design and performance and sharpened my programming and data analysis skills. Overall, the internship was challenging and rewarding, providing me with invaluable knowledge and confidence in my abilities.
Battery State of Health Estimation and Cooling Circuit
Research Project, BITS Pilani
Pilani, India
August '21 - December '21
In collaboration with a peer, I spearheaded the development of an innovative system to regulate battery temperature and performance in electric vehicles as part of an independent research project. This experience deepened my knowledge of battery chemistry, design, and thermal properties. Together, we built a detailed simulation model using Simulink and MATLAB, consisting of six core components and several additional modules. To enhance its realism and accuracy, we incorporated random Gaussian noise, making the model a robust tool for testing and analysis.
Rarefied Gas Flow Analysis in 2D Annulus Ring
Gas Dynamics Course Project, BITS Pilani
Pilani, India
November '21 - December '21
I undertook a challenging project in Gas Dynamics, alongwith another student of the course, analyzing rarefied gas flow in a 2D annulus ring. Through advanced CFD simulations using kinetic theory of gases and Boltzmann equations, we presented key insights on the impact of different annulus ring speeds and directions on gas flow. We performed our simulations using ANSYS Fluent.
Inspired Karters - FSAE Race Car Team
Aerodynamics Head
Pilani, India
September '19 - May '22
I have been an F1 fan all my life and was a part of an electric racecar team. I contributed to the Aerodynamics subsystem majorly, working on design, analysis and optimisation of wings, undertray and diffusers.
We were able to achieve 4 second faster lap times through aerodynamic improvements and bagged 5 internationally acclaimed recognitions throughout my 3 years being part of the team.
Inspired Karters indeed brought together the smartest and the most passionate Motorsport lovers who are turning out to be excellent engineers. I equally enjoyed my CFD analysis and workshop fabrication sessions!
Hyperloop India
Propulsion Subsystem
Remote
August '21 - October '21
I was briefly a part of the Hyperloop India team which was based out of BITS Hyd but had members from all the top Indian engineering institutes. I was part of the propulsion subsystem and worked on simulating electric and magnetic fields inside induction motors to improve performance. I had to discontinue it owing to the pandemic but I worked with some very intelligent people and a short but worthwhile learning experience.
Volunteering
Shiksha ki Ore - Nirmaan, BITS Pilani Chapter
Community Development Volunteer
Pilani, India
August '19 - May '23
My 4 year NGO experience, where we worked for the development of 25+ families was magical to say the least. My team comprised of 20 young change-makers and tackled complex community issues like child marriage, child labour, education and sanitation. I was entrusted with leading the project for a semester during the pandemic wherein the team was able to ensure ZERO covid cases in the community. I have met the most kind and hardworking individuals here whom I am grateful to call friends!