Shivam Vashi

About Shivam Vashi

I am a passionate and curiosity-driven mechanical engineering graduate student at Texas A&M University pursuing my Master of Science in Mechanical Engineering. My academic journey has been guided by an insatiable curiosity about autonomous vehicles motion planning, vehicle dynamics & controls and perception, and the complexities of UAVs and UGVs.My skills include programming in Python, MATLAB, C/C++, and Machine Learning, as well as working with software such as ROS, Gazebo, MATLAB/Simulink, ANSYS FEM & CFD, and Fusion 360.As a Graduate Autonomous Vehicle Research Assistant, I tested autonomous driving algorithms in both controlled environments and on public roads, including rural highways. I designed a ROS-based Planning pipeline integrating Object Tracking, Trajectory Prediction, and Control Reference Generation, ensuring smooth communication with Perception and Control modules. I implemented decision-making for high-speed maneuvers (45 mph) and real-time GPS trajectory adjustment using Meta SAM2 to eventually allow GPS-denied Path Planning (which I\'m currently working on). Additionally, I developed emergency stop strategies, applied Lidar for road width estimation, and enhanced sensor fusion output with Kalman Filters to improve object tracking.On my previous assignment as a Graduate Student Assistant woring on quantifying and reducing Simulation-to-Realtiy Gap, I implemented and deployed A* path planning algorithm and Pure Pursuit controller on Clearpath Husky robot and its digital twin in Gazebo. I also worked on closing the simulation-to-reality gap using Hardware-in-the-loop Testing and System Identification.In addition to my academic and professional journey, I am actively involved in hands-on projects. I\'ve worked on re-implementing a Nonlinear Model Predictive Control strategy for autonomous racing of scale vehicles. I\'ve also programmed a self-driving vehicle simulation, which involved autonomous navigation using longitudinal (PID) and lateral (Stanley) controllers.My dedication extends to extracurricular projects, such as designing and building an RC aircraft during my tenure as the Team Leader and General Secretary for collegiate Aeromodelling club- Aero VJTI along with developing an autonomous line-following and self-balancing unicycle robot during a workshop organized by SRA VJTI Robotics ClubI am eager to collaborate, learn, and contribute to projects and teams that share my enthusiasm for Self-Driving Cars and Robotics. Let\'s connect!

Experience

1. Motion Planning Software Engineer

   Woven by Toyota

   Sep 2025 — Present · Ann Arbor, MI, US

   Contributed to Guardrail Architecture (downstream of ML Planner) for lane-change safety, including a Stanley-controller-based fallback planner for lane-change abort, TTC/critical-gap utilities, and a mock Behavior Proposer builder for robust testing- Led guardrail visualization enablement by defining visualization requirements, wiring message format converters, and ensuring key guardrail signals (constraints, lane-change status, turn indicators) are available in simulation, missions, and triage tools- Drove CppTest static analysis adoption for production motion-planning code, integrating it with Bazel targets and enabled broad cleanup of MISRA violations using GitHub Copilot–assisted fixes.

Education

• Veermata Jijabai Technological Institute (VJTI)

  Bachelor of Technology - BTech, Mechanical Engineering

  2019 — 2023

• Texas A&M University

  Master of Science - MS, Mechanical Engineering