Military academy honored in GPS-free drone competition in Chiayi

Taipei, April 29 (CNA) A major military academy in Taiwan has won a prize at a university-level drone competition that required participants to perform drone reconnaissance missions without using satellite navigation systems, to test their operational skills, according to a Ministry of National Defense (MND) press conference Tuesday.

The second and final round of the "2025 UAV Defense Challenge" took place last month at the Asia UAV AI Innovation Application R&D Center in Chiayi County. The competition was based on a scenario where drones are being jammed and cannot link to the Global Navigation Satellite System, forcing operators to fly drones independently while conducting reconnaissance missions, locating their targets and determining their numbers, according to the event's website.

The competition was hosted by National Cheng Kung University and overseen by the National Science and Technology Council. The winners were announced on April 11.

A team from National Defense University's (NDU) Chung Cheng Institute defeated 14 teams and was among the competition's six finalists. The jury presented it with a "best potential" award.

The competition's theme was inspired by rescue missions in the aftermath of the Hualien Earthquake on April 3 last year when rescue missions in Taroko Gorge were made more challenging by the lack of satellite navigation, Major General Cui Yi-feng (崔怡楓), head of the institute, told the MND news conference.

Using a commercially available drone, the NDU team finished the competition using "dead reckoning" - the process of calculating the current position of a moving object by using a previously determined position and incorporating estimates of speed, direction, and elapsed time, Cui said.

To ensure the drone was up to the task set out by the competition, the team outfitted it with an arsenal of cutting-edge technologies, including an Nvidia Jetson Nano computing board, a Global Shutter image sensor, and a deep-learning inference optimizer, to allow it to recognize the contest's predetermined targets, including civilian vehicles, camouflaged military vehicles, and oil barrels, Cui explained.

The NDU institute will share its findings from the competition with Taiwan's government-owned weapons developer National Chung-Shan Institute of Science and Technology and private drone manufacturers to facilitate research and development and bring them more in line with the military's needs, Cui said.

Despite the competition having a civilian defense theme, Shu Hsiao-huang (舒孝煌), a research fellow at the Institute for National Defense and Security Research, a think tank, told CNA that applications in the competition are also useful in wartime scenarios, for example, when an enemy jams satellite navigation.

"Whether it is topology-related restrictions or the result of jamming that causes drones to be unable to receive signals, the same principles apply," Shu said.

As the ongoing Russia-Ukraine War has demonstrated, the use of electronic warfare to jam drones is a very potent and cost-effective tactic, Shu noted.

To counter jamming, AI-powered modes of drone operation were developed as the war progressed, in response to a scenario when satellite navigation is not available, Shu explained.