VEX Robotics is a yearly world-wide competition on levels ranging from middle school to college. VEX introduced an AI competition in 2021, so I joined Purdue’s newely formed VEXAI team as a member of the reinforcement learning subteam. The competition never happened due to supply chain issues, nonetheless our team was able to build an autonomous robot that could navigate through the playing field and pick up rings (the games objective). You can see our technical report at Pac-Man Pete An extensible framework for building AI in VEX Robotics.

Role

2021-2022: Member - Reinforcement Sub-team

2022-2023: Project Lead

Tech Stack

• Reinforcement Learning
  • Unity 3D Game Engine
  • Unity Machine Learnings Agents Package
  • C#
  • Linux Server
• Computer Vision
  • Python
  • Pytorch
  • Yolov5
  • Roboflow
• Hardware/Robotics
  • VEX Brain
  • PROS (PROS Robotic Operating System)
  • C/C++

Pipeline

Results

This was my first project I got on at Purdue and helped me meet a bunch of great people in the machine learning space as well as force me into building a project with machine learning instead of tutorials or classes like I was used to. If you noticed my sophomore year I became the head of the VEXAI team where we built out another computer vision system for the robot. After my sophomore year I left the project just to expand my range of projects and go into another venture. Producing a technical report, leading a project, and incorporting a wide stack with multiple subteams taught me a lot and set me up for my next project AI Maritime Manuever Challenge.

---