We are excited about the critical and creative thinking about the AlphaPilot qualifications and competition!

We would like to update all the teams on some technical details for the AlphaPilot competition to help your team formulate an approach and answer Test #1 questions.

All teams that qualify for the AlphaPilot competition will utilize the same drone configuration. Here are some highlights about the DRL AIRR quadcopter technical specifications:

Sensor Suite

  • 3 x stereo camera pairs provide over 180 deg field-of-view and are the primary feedback source for navigation
  • 6-DOF IMU with gyroscope, accelerometer, and magnometer provide odometry
  • Incorporation of a laser rangefinder

Computational Resources

  • NVIDIA Tegra Xavier GPU will provide computational power to the drone
  • Utilization of NVIDIA Deep Learning SDK, NVIDIA VisionWorks, CUDA Toolkit, TensorFlow, OpenCV
  • Teams will operate the AIRR drone in ACRO/RATE mode where the input is angular velocities

Development and Testing

  • Each qualifying team will receive a Development Kit consisting of the AIRR Rule Book, development program including coding tools and access to the DRL AIRR Simulator, and a skeleton version of the DRL AIRR drone (i.e. a platform with the same sensing and computational resources but without a propulsion system) which can be used for development and testing
  • The primary resource available to teams for development and testing will be the Development Kitwhere all programming will be in C++/CUDA
  • LMCO and DRL are planning to host workshops and other opportunities for teams to share best practices, meet with their Lockheed Martin team mentors, and discuss development with engineers from AlphaPilot and AIRR