Shubham Parab

About

Robotics Engineer with 3+ years building production autonomous systems, specializing in ROS 2 control stacks, MoveIt2 manipulation, and SLAM navigation. Currently leading perception and localisation efforts for Formula Student AI race cars.

Experience

• Research Assistant – 3D Vision
  Buckinghamshire New University · High Wycombe
  Mar 2026 – Jun 2026

  Developed an underwater 3D reconstruction system integrating 3D Gaussian Splatting (3DGS) with a lightweight neural network to model light absorption, scattering, and colour distortion. Reduced training time by over 90% versus prior NeRF-based methods, achieving real-time rendering at up to 241 FPS.

• Research Intern – Computer Vision
  Newcastle University · Newcastle
  Mar 2026 – May 2026

  Built a 2-stage CV pipeline using SAM3 for zero-shot object detection and a fine-tuned ConvNeXt CNN for site suitability classification, achieving 98% accuracy. Engineered a geospatial Street View scraping workflow cutting manual planning workload by 97.5%.

• Systems Engineer – AI & Robotics
  Tata Consultancy Services · Bengaluru
  Jun 2022 – Aug 2025

  Architected ROS 2 control and motion stacks for production robot platforms. Implemented MoveIt2 manipulation pipelines for UR5 robotic arm. Engineered autonomous navigation stacks integrating SLAM and Nav2. Developed 6D pose estimation and retail shelf anomaly detection pipelines (PyTorch, OpenCV).

Education

• MSc Data Science & Artificial Intelligence
  Newcastle University
  Data Science & Artificial Intelligence · Sep 2025 – Sep 2026
• BE Electronics Engineering
  RAIT, Mumbai University · Mumbai University
  Electronics Engineering · 2018 – May 2022

Skills

• Computer Vision
• SLAM
• ROS 2