Getting Started Guide - UAV Mission Compute SDK#

Overview#

The UAV Mission Compute SDK provides a comprehensive development environment for UAV (Uncrewed Aerial Vehicle) applications using Intel’s optimized compute tools and frameworks. It packages a PX4 + Gazebo simulation with multi-camera support, OpenVINO-based vision processing on Intel GPU, MQTT telemetry, and RTSP streaming — all orchestrated via Docker Compose.

Learning Objectives#

Upon completion of this guide, you will be able to:

  • Install and configure the UAV Mission Compute SDK

  • Launch the PX4 simulation stack with simulated cameras

  • Start the AI vision processing pipeline

  • View live RTSP camera streams with Intel Edge AI inference

System Requirements#

Verify that your development environment meets the following specifications:

  • Operating System: Ubuntu 24.04 LTS (provisioned using Edge-Node Infrastructure Blueprint)

  • Memory: Minimum 16GB RAM (32GB recommended)

  • Storage: 100GB available disk space

  • Network: Active internet connection for package downloads

  • Hardware: Intel Core Ultra Series 3 (Panther Lake) processor with integrated GPU recommended

Installation Process#

Execute the automated installation script to configure the complete development environment:

curl -fsS https://raw.githubusercontent.com/open-edge-platform/edge-ai-suites/refs/heads/main/metro-ai-suite/metro-sdk-manager/scripts/uav-mission-compute-sdk.sh | bash

UAV Mission Compute SDK Installation

The installation process configures the following components:

  • Docker containerization platform

  • PX4 autopilot simulation with Gazebo Harmonic

  • Multi-camera bridge (nadir, forward, rear at 416×416 @20fps)

  • Companion telemetry bridge (MAVLink → MQTT)

  • MQTT broker and MediaMTX RTSP server

  • InfluxDB time-series storage and Grafana dashboards

  • Metrics manager for host platform monitoring

  • OpenVINO-based vision processor (YOLOv2 on Intel GPU)

Once the script completes, the full stack is built and running:

UAV Mission Compute SDK Completion

UAV Mission Compute SDK Application Setup#

This section describes how to verify and interact with the running application.

Step 1: Wait for PX4 to Be Healthy#

The installation script starts the simulation stack automatically. Wait for PX4 to become healthy (~60–90 seconds on first boot):

cd ~/oep/edge-ai-suites/federal-and-aerospace-ai-suite/uav-mission-compute-sdk
docker compose ps px4

Step 2: Arm the UAV (Activate Cameras)#

Cameras only stream when the UAV is armed. Arm it via the REST API:

curl -X POST http://localhost:8080/action/arm

Step 3: View RTSP Streams (Optional)#

View any camera feed using an RTSP player:

ffplay rtsp://localhost:8554/uav-1/nadir

Step 4: Stop the Application#

To stop the entire infrastructure stack:

make down

Technology Framework Overview#

UAV Mission Compute SDK Components#

The UAV Mission Compute SDK integrates multiple technologies:

  • PX4 + Gazebo Harmonic: Flight controller simulation with multi-camera world

  • Companion Bridge: MAVLink ↔ MQTT telemetry bridge

  • Camera Bridge: Gazebo frames → H264 → RTSP via MediaMTX

  • OpenVINO Vision Processor: Real-time YOLOv2 vehicle detection on Intel GPU

  • MQTT Broker (Mosquitto): Lightweight messaging for telemetry and detections

  • InfluxDB + Grafana: Time-series storage and dashboards for flight and platform metrics

Next Steps#

After installation completes:

  1. Navigate to $HOME/oep/edge-ai-suites/federal-and-aerospace-ai-suite/uav-mission-compute-sdk/ to explore the SDK

  2. Review the Get Started section for USB camera setup and advanced configuration

  3. Access Grafana dashboards at http://localhost:3000 (admin/admin)

  4. Explore the REST API at http://localhost:8080 for flight control commands

Additional Resources#

Technical Documentation#

  • OpenVINO - Intel’s cross-platform inference optimization toolkit

  • Edge AI Libraries - Comprehensive development toolkit documentation and API references

  • Edge AI Suites - Complete application suite documentation with implementation examples

Support Channels#