Arthur Garzon

About Me

MISSION

Engineer unmanned aerial systems that keep America ahead.

Experience

INTERNSHIPS

AI/ML Engineering Intern

Zone 5 Technologies

July 2025 - September 2025

San Luis Obispo, California

  • •Collaborated under the Head of AI/ML to lead integration and debugging of sensor payload for Paladin, a Group 2 UAS, enabling flight and Counter-Unmanned Aerial System (C-UAS) capabilities in GPS-denied environments.
  • •Developed Python scripts using fastdup to deduplicate YOLO datasets by 10%, streamlining dataset preparation, and Segment Anything Model 2 (SAM2) to convert bounding box annotations to polygons, improving segmentation model training efficiency.
  • •Fine-tuned YOLOv8 segmentation model on over 60,000 UAS images, setting the standard for Paladin's vision system and improving target localization by 9.5%.
  • •Updated Paladin's vision codebase to support YOLO segmentation models on Nvidia Jetson hardware.
  • •Managed GPS-denied payload operations via SSH-based Linux Command Line Interface, ensuring reliable performance during tests and demonstration at Camp Roberts Base.

PROJECTS

Team Lead - Fire-mapping VTOL Drone

UCSB Engineering Capstone Program

September 2025 - Present
  • •Won the CNSI Innovation grant and awarded $3,000 in funding to create and lead engineering capstone team.

Counter-UAS Payload

Independent Project

August 2025 - Present
  • •Developing a modular payload for PX4/ArduPilot-based FPV drones, enabling kinetic interception for C-UAS operations.
  • •Prototyped UAS detection using a fine-tuned YOLOv11n model (mAP50 over 80%) deployed on Raspberry Pi and Nvidia Jetson Orin Nano hardware.
  • •Validated market needs through discussions with defense officials, including a U.S. Army Colonel and the small Unmanned Aerial System (sUAS) Division Chief, Army Capability Manager-Unmanned Aircraft Systems (ACM-UAS).

Real-Time Drone Meeting Platform

Independent Collaboration

March 2025 - April 2025

Santa Barbara, CA

  • •Built a web app using livekit-server-sdk and Web Real-Time Communication (WebRTC) for low-latency video streaming from drone platforms, enabling real-time monitoring.
  • •Integrated the web app with PencilSpaces to support virtual meeting participation via drone video feeds.
  • •Hosted an HTML frontend with control buttons via ngrok on a Raspberry Pi, providing user-friendly drone control.
  • •Pitched the prototype to a venture capitalist, receiving positive feedback on its potential for remote surveillance applications.

Home Surveillance Drone

Independent Project

July 2024 - February 2025
  • •Designed and 3D-printed two indoor drone prototypes, performing FEA using Solidworks and flight tests to validate structural integrity.
  • •Developed Python scripts using pymavlink to interface a Raspberry Pi onboard computer with a PX4 flight controller, enabling autonomous patrols for indoor surveillance.
  • •Built a Flask web app to serve as a user-friendly ground control station for drone monitoring and control.