To make this website work, we log user data. By using Shephard's online services, you agree to our Privacy Policy, including cookie policy.

×
Open menu Search

General Atomics flies UAS missions with AI pilots

13th January 2023 - 16:00 GMT | by The Shephard News Team in London

RSS

GA-ASI's flights took advantage of its reinforcement learning architecture. (Photo: GA-ASI)

A GA-ASI-owned Avenger UAS was paired with a digital twin aircraft to conduct live, virtual, and constructive combat missions autonomously.

General Atomic Aeronautical Systems (GA-ASI) has advanced its Collaborative Combat Aircraft (CCA) ecosystem by flying three missions with artificially intelligent (AI) pilots.

The flights took place on 14 December from GA-ASI’s Desert Horizons flight operations facility as a demonstration of its commitment to maturing the CCA ecosystem for Autonomous Collaborative Platform (ACP) UAS using AI and Machine Learning (ML).

The flights took advantage of GA-ASI’s reinforcement learning (RL) architecture, with agents demonstrating single, multi and hierarchical behaviours.

The single-agent model navigated a live aircraft while avoiding threats. Multi-agent RL models flew a live and virtual Avenger UAS collaboratively to chase a target and avoid threats.

Finally, the hierarchical RL agent used sensor information to select courses of action based of its understanding of the environment.

GA-ASI said this demonstrated the AI pilot’s ability to process and act on real-time information independently of a human operator.

MQ-20 Avenger

The Shephard News Team

Author

The Shephard News Team


As part of our promise to deliver comprehensive coverage to Premium News and Defence Insight …

Read full bio

Share to

Linkedin