AIRHOUND Research Project

PX4 SITL HITL ROS2 LaTeX

Overview

AIRHOUND (Autonomous Intelligent Rotorcraft for Hostile Unidentified Navigation and Detection) is an 8-person research project at Embry-Riddle Aeronautical University's Engineering Physics Propulsion Lab. The project's goal is to detect and track airborne objects in real time, with applications in aerospace, military, and naval domains. The system runs on an NVIDIA Jetson Orin using the ROS2 Humble uXRCE-DDS framework.

My Contribution

On this project, I helped integrate the PX4 converter node into our drone framework, which let the flight controller and onboard Jetson Orin compute communicate reliably over ROS2's uXRCE-DDS middleware. I also configured our Software-in-the-Loop (SIL) and Hardware-in-the-Loop (HIL) testing pipelines, which is a critical step for catching issues in simulation before they ever reached a live flight test. Alongside this, I co-authored Section 2 (Methods) for our team's paper, which was published in an undergraduate research journal.

Gallery

Outcome

Published in ERAU Beyond: Undergraduate Research Journal, Vol. 9 (2025). The paper details our ROS2 integration approach and testing methodology and is available through ERAU Scholarly Commons.

View Paper

What I Learned

AIRHOUND taught me how tightly software and hardware have to work together in a real-time system, where small timing issues between the flight controller and onboard compute can become real problems. Running SIL and HIL tests reinforced the value of catching failures in simulation rather than in the field. It also sharpened my technical writing skills.