Defense scientists from Australia, the United Kingdom and the United States collaboratively tested a range of robotic vehicles and their sensors during a trial at the Cultana training area in South Australia.
The Reliable Operation of Robotic Vehicles in a Contested Environment (TORVICE) test was conducted in fall 2023 to identify and resolve vulnerabilities that autonomous systems face in a congested electronic warfare environment.
Dr Peter Shoubridge, Head of Defense Joint and Land Warfare, said the test was designed to test the behavior of autonomous vehicles when under attack.
"Understanding how robotic vehicles react in challenging environments accelerates our collective knowledge and helps improve the system to overcome such attacks," said Dr. Shoubridge.
A network of UK and US robotic ground vehicles was configured to represent autonomous multi-domain launchers and unmanned ground vehicles carrying out long-range precision fires and associated missions. The vehicles did not carry weapons during the trial.
Australian scientists then subjected the vehicles to electronic warfare, electro-optical and position, navigation and timing attacks.
"Transitioning reliable robotic capabilities into the hands of our warfighters safely and ethically is a priority," said Dr. Shoubridge.
TORVICE is part of the United States, United Kingdom and Australia's commitment to the AUKUS Advanced Capabilities Pillar, known as Pillar II, under the Artificial Intelligence and Autonomy Working Group. The test is part of an ongoing series pursuing a trilateral work program on a range of cutting-edge technologies and capabilities to promote security and stability in the Indo-Pacific region.
Dr. Kimberly Sablon, Senior Director of Trusted Artificial Intelligence and Autonomy at the US Department of Defense, said: "The TORVICE project builds on the work AUKUS partners demonstrated in Salisbury in April. During this exercise, "We conduct rigorous red teaming of our autonomous/AI systems to assess and mitigate vulnerabilities and improve their resilience in complex and contested environments."
Through AUKUS, Australia, the United Kingdom and the United States have collaborated to accelerate the collective understanding of AI and autonomy technologies, and how to rapidly deploy robust and reliable AI and autonomy in complex operations, while adhering to the shared values โโof safe and responsible AI. .
TORVICE follows the first AUKUS autonomy and artificial intelligence (AI) trial held in the UK in April. The goal of this AUKUS collaboration is to rapidly advance these technologies towards responsible military use.
