Australia improving rescue efforts with artificial intelligence
The Royal Australian Air Force (RAAF) is harnessing the power of artificial intelligence (AI) to enhance its search-and-rescue (SAR) capability in a region where earthquakes, tsunamis, floods and tropical cyclones claim an average of over 40,000 lives each year.
“Our vision was to give any aircraft and other defense platforms, including unmanned aerial systems, a low-cost, improvised SAR capability,” Wing Commander Michael Gan, who leads AI development for RAAF’s Plan Jericho, said in a news release from Australia’s Department of Defence. Plan Jericho, which was launched in 2015, is an RAAF 10-year blueprint to become one of the world’s most technologically advanced air forces.
Christened AI-Search, the project involves “training a machine-learning algorithm and AI sensors to complement existing visual search techniques,” Gan said.
AI-Search was successfully tested for the first time in October 2019 aboard an RAAF C-27J Spartan aircraft, pictured making a low pass to detect floating objects below, the release stated. The plane flew the AI-Search team and sensors over the Tamar River in Tasmania while the system learned to identify a range of life rafts, lifeboats and safety equipment.
The second trial took place near Stradbroke Island in Australia’s Queensland province in March 2020. This time, AI-Search detected several small targets in a wide sea area.
AI-Search began as a challenge from RAAF Air Commodore Darren Goldie, director general of air combat capability, to improve SAR using enhanced sensors. It eventually combined the resources of Plan Jericho with contributions from the RAAF Air Mobility Group’s No. 35 Squadron, the Royal Australian Navy’s Warfare Innovation Branch and the University of Tasmania’s Australian Maritime College.
“There is a lot of discussion about AI in [the Department of] Defence, but the sheer processing power of machine learning applied to SAR has the potential to save lives and transform SAR,” Lt. Harry Hubbert of the Navy’s Warfare Innovation Branch, who developed algorithms for AI-Search, said in the news release.
Previously, Hubbert worked on developing AI-enabled autonomous maritime vehicles for the collaborative Autonomous Warrior exercise, supported by the Royal Australian Navy, in Australia’s Jervis Bay in 2018.
AI-Search was configured using largely off-the-shelf commercial components with custom software and programming by Hubbert and his team. So far, the trials have highlighted the feasibility of the technology, which Australia’s Department of Defence reported can be applied easily to other airborne platforms.
RAAF Capt. Jerome Reid credited the project’s success to its free-flowing collaborative character. “Empower our bright sparks, connect them with mentors and resources, get out of their way and let the magic happen,” he said in the news release.
Tom Abke is a FORUM contributor reporting from Singapore.