GitHub Repository

An open framework to simulate and deploy perception-based PX4/ArduPilot drone swarms with ROS2, YOLO, LiDAR, NVIDIA Jetson

500 starsC++

Aerial-autonomy-stack – open-source perception-based drone swarms

Name: Aerial-autonomy-stack – open-source perception-based drone swarms
Availability: InStock
Author: SufficientFix42

by SufficientFix42·Jun 27, 2026·2 points·0 comments

Visit Project View on HN

AI Analysis

●●●BangerWizardryBig BrainNiche Gem

One-liner 45-minute build for full drone autonomy stack is genuinely impressive.

Strengths

•Collapses simulation, CI/CD, and deployment into single containerized workflow
•KISS-ICP LiDAR odometry and Zenoh inter-vehicle ROS2 bridge
•Multi-Jetson HITL simulation for NVIDIA ARM-based compute testing

Weaknesses

•Ubuntu and NVIDIA driver requirements limit accessibility
•Niche robotics audience means limited broader adoption

Post Description

Combining (1) open-source autopilots with (2) open-source ROS2 simulation/navigation/mapping packages and (3) open-source, hardware-accelerated computer vision should not be hard, but it is.

aerial-autonomy-stack aims to solve this by collapsing (1) multi-robot, (2) software-in-the-loop, (3) hardware-in-the-loop simulation, (4) CI/CD, and (5) deployment into a single containerized workflow.

The end-to-end stack can be re-created/simulated with a one-liner (45') build on Ubuntu >=22.