Perception & Computer Vision | AI/ML | ROS2 | OpenRMF | SLAM | Autonomous Systems | Embedded Systems

Education

University of Maryland, College Park. M.Eng. in Robotics 2023 – 2025

Pune University B.Eng. in Electronics and Telecommunication 2019 – 2023

Aug 2024 - Present

Conducted system diagnostics and troubleshooting for a fleet of autonomous forklifts, implementing real-time monitoring tools that reduced system failures by 20%.
Partnered with R&D and systems teams to integrate RoboShop (SEER Robotics) for real-time management of autonomous forklifts and AMRs in dynamic warehouse setups.
Conducted on-site diagnostics and training, enhancing client satisfaction and improving delivery timelines and reducing operational downtime by 20%.
Collaborated with R&D teams and senior stakeholders to ensure timely deployment, workflow coordination, and on-site issue resolution.

Oct 2021 - Dec 2021

Collaborated closely with industry experts to optimize business processes and gained expertise in multiple projects leveraging a diverse range of technologies, including Arduino, Jetson Nano, Jetson Xavier, and Autodesk Fusion 360.
Implemented Machine Vision cameras across diverse domains, streamlining quality control processes and reducing defects through automated visual inspection.
Spearheaded the development of a Laser Gauge Micrometer, effectively communicating complex concepts to colleagues and senior management, resulting in increased buy-in and project alignment.

May 2021 - Aug 2021

Analyzed experimental data with precision and proactivity, translating findings engaging presentations leading to end product development; showcased the potential of UV sanitation for improved health and safety.
Led system redesign initiative, streamlining processes and optimizing workflows, resulting in a 30% increase in team efficiency.
Rendered CAD designs to optimize 3D models for UV-based sanitation solutions.

Oct 2024 - Dec 2024

Led the development of a Language Model trained on Sanskrit texts from the Atharva Veda.
Achieved near-optimal perplexity scores of 1.02 by fine-tuning the MuRIL model on an custom dataset (KANDA-1), demonstrating exceptional performance in Sanskrit text processing.
Implemented dynamic masking strategies and a custom training approach, resulting in validation loss reduction from 0.13 to 0.024 during the initial fine-tuning phase.
Compared MuRIL with adapter layers, improving model efficiency by reducing the number of trainable parameters while still achieving a validation loss of 0.840.

Mar 2024 - Mar 2024

Implemented video processing pipeline using OpenCV for edge detection, Hough Transform, and corner detection. Developed image stitching pipeline using feature extraction, RANSAC for feature matching, and homography computation.
Achieved 95% accuracy in detecting and highlighting paper corners in videos, removing 20% of blurry frames. Created seamless panoramic images with over 87% feature match accuracy, enhancing image clarity and consistency.

Apr 2024 - Apr 2024

Developed an OpenCV-based calibration pipeline and stereo vision system for accurate corner detection, parameter optimization, and depth map computation.
Achieved sub-pixel accuracy with 0.3 pixel reprojection error and 80% depth map precision, enhancing 3D spatial understanding.

Nov 2023 - Dec 2023

Employed a 5 DOF manipulator arm with a claw-like end-effector for intricate handling.
Developed and optimized precise forward kinematics models for manipulator arms, ensuring accurate end-effector positioning and achieving a 95% reduction in positioning errors.
Validated kinematic calculations with exceptional accuracy and efficiency through meticulous testing methodologies; reduced margin of error by 75%.
Optimized the tele-operation model using Python, resulting in a 12% increase in handling time and movement.

Dec 2023 - Dec 2023

Built ROS2 package for autonomous TurtleBot navigation in intricate mazes.
Built a project utilizing Aruco markers, significantly improving localization precision and reducing collision incidents by 12%, resulting in enhanced performance for autonomous vehicles.
Orchestrated the integration of Python and C++ for real-time object recognition, enabling advanced environment analysis techniques.
Enhanced communication using ROS nodes, enabling faster response times.

Jun 2022 - Jan 2023

Enhanced embedded control systems for an electric vehicle, increasing energy efficiency by 15%.
Developed a real-time data acquisition and analysis system to optimize vehicle performance.
Designed a user-friendly interface for vehicle control, increasing user engagement by 20%.

Languages: Python, C++, Embedded C, Git, Scripting (Bash), Linux CLI, LaTeX, HTML
Libraries: NumPy, Sympy, Keras, TensorFlow, Pandas, Matplotlib, OpenCV
Robotics: ROS 2, RViz, Gazebo, OpenRMF
Development Boards: Arduino, AtMega, Nvidia Jetson, Raspberry Pi, Teensy, Orange Boards

Online Resume for Suraj Kalwaghe.