Home / Qualcomm® QRB5165 Drone - Build Autonomous Drones using Qualcomm Processors

Build Autonomous Drones with Qualcomm® Processors with ModalAI®

Qualcomm® Snapdragon™ processors enable developers to create the smallest autonomous drones out there. ModalAI's VOXL 2 is an improved version of the Qualcomm® Flight Pro architecture, and includes comprehensive documentation and support by the same industry leading engineers that developed Snapdragon Flight and Qualcomm Flight Pro at Qualcomm. Accessibility and ease of use are really important to the team at ModalAI, and much of their effort is spent making autonomous drones easy to develop and deploy for a broad range of applications.

Open Development Drone with Qualcomm Snapdragon

VOXL 2 improves on Qualcomm Flight Pro and Snapdragon Flight by including: 
  • PX4 Flight Control, Avoidance and GPS-denied Navigation (VIO)
  • ROS and MAVROS Autonomous Navigation
  • MAVSDK support on board for building new applications
  • Snapdragon Navigator flight controller
  • Beyond Visual Line of Sight flight over 4G / LTE (demo)

Ready to Use Kits Accelerate Development

We make many configurations of VOXL available to ease integration into your sUAS or robotics application:

  • Stand-alone Development Kit (Buy Here)
  • Machine Vision Development Kit (Buy Here)
  • Mounted and calibrated Development Kit (Flight Deck Buy Here)
  • A complete Reference Drone! (m500 Buy Here)
  • A Snapdragon and PX4 flight controller in one PCB (VOXL Flight Buy Here)

 

 

Software Features

  • Open source Linux kernel, cross-compilers, PX4, ROS, OpenCV
  • Docker build environment for CPU, GPU (OpenCL) and DSP (Hexagon SDK) heterogeneous computer vision and deep learning processing.
  • Run Docker on VOXL target for improved portability and maintenance.
  • Add cellular LTE connectivity for beyond visual line of sight (add-on).

Specifications

  • Snapdragon 821: Quad-core up to 2.15GHz, GPU, 2xDSP, Linux
  • Video support: 3x MIPI cameras, UVC 
  • Machine vision support: time synchronized IMU and cameras
  • Additional I/O: WiFi, SPI, I2C, UART, GPIO
  • Light weight, powerful, SWAP-optimized computing:
    • PCB weight 18g
    • Power Module weight 13g
    • Cameras (6g)
    • Power consumption 3-10W