As North Korea ramps up suicide drone production, South Korea hits a wall at sea

BUSAN — For South Korea's drone warfare, the sky isn't the limit. The sea is. Over the waters off the country's southern coast on Tuesday, a domestically developed long-range loitering munition pitched sharply upward and unexpectedly crashed into the sea seconds after launch. About two hours later, a second prototype followed suit. With North Korea having acquired similar technology from Russia and begun mass production of its own loitering munitions, the South Korean military has effectively encountered the limitations of its technology.

Reporters watched both attempts from the bow of the ROKS Marado, the Navy's 14,500-ton amphibious assault ship, in international waters south of Busan. The Agency for Defense Development (ADD) and the Navy had invited the press to the drone's first maritime operational trial, the final hurdle before system development is due to finish in September.

The aircraft is the basic model of K-Lucas, or K-Low-cost Uncrewed Combat Attack System, a domestically developed long-range loitering munition. The ADD began developing its core technologies in June 2024 under an early-stage defense research project shaped by lessons from the war in Ukraine. A loitering munition can remain airborne while searching for a target before diving into it and detonating.

The Navy wanted the drone to find and strike a high-speed unmanned target boat floating off the Marado's starboard side, and a second drone to follow it to conduct a battle damage assessment determining whether the target had been destroyed.

An ADD official remotely started the drone on the forward deck. Its engine broke into a loud rattle not unlike the distinctive sound of Iran's Shahed-136 attack drone. Shahed drones' characteristic din has now become familiar to ears across Ukraine after years of Russian attacks. Russia operates large numbers of modified Shahed-derived drones.

The South Korean prototype also looked much like a Shahed. It was about 3 meters (9.8 feet) long, weighed 150 to 200 kilograms (330 to 440 pounds) and carried an explosive payload of several dozen kilograms. It launched from a wheeled launcher secured to the bow, where two rectangular canisters were stacked vertically so two drones could be fired in succession.

At 3:30 p.m. a burst of rocket flame shot from the first drone's tail as it left the launcher and turned toward the target boat. Within one or two seconds, however, its nose pitched sharply upward. The rear propulsion unit produced an unstable burst of flame before the aircraft plunged into the sea. The entire flight lasted only seconds.

Thick, acrid smoke then spread across the deck and briefly obscured the view. For a moment, the scene resembled an actual combat engagement.

The ADD tried again about two hours later. The second drone pitched upward after three to four seconds and crashed into the sea in nearly the same way. The Navy and the ADD could not evaluate any of what they had come to test, including target detection, the strike itself or the damage assessment.

Combat tests that the military opens to reporters do not usually fail outright. Afterward the Navy said it would not release the photographs and video it had promised. The ADD had been confident enough to run no rehearsal, and for reasons still unknown the drone never reached normal flight, let alone struck anything.

Sea fog was light on Tuesday and the wind and waves were relatively steady.

"The engine and other key parameters were normal, and the system was operating normally," an ADD official said. "We have flown without problems on land in stronger winds than these. We plan to conduct a detailed investigation into the cause based on the data collected."

The ADD will examine whether a minor defect developed when the drone and launcher were disassembled for transport aboard the ship and then reassembled.

The Navy and the ADD say the unsuccessful trial does not mean the underlying technology has failed because testing is part of the development process. The ADD says the system has completed around 30 land-based sorties and carried out missions against fixed and moving targets. The maritime trial was meant to clear the last major hurdle before development wraps up next month.

Whether the drone can operate at sea, however, is what will determine its practical effectiveness. Military experts say North Korea could try to slip loitering munitions in covertly over the East and Yellow Sea in a conflict, or fly them in swarms, because open water presents fewer obstacles than land and suits a drone traveling a long distance. The sea could also provide South Korea with a route for drone counterstrikes. A system that cannot launch and fly reliably over water could therefore offer only partial capability.

North Korea has been pushing its own one-way attack drones toward mass production as it seeks an asymmetric edge over South Korea. North Korean leader Kim Jong-un ordered officials in November 2024 to build a serial production system as quickly as possible and begin full-scale mass production of what state media calls "suicide attack drones."

North Korea has also deepened drone cooperation with Russia. Kyrylo Budanov, head of Ukraine's Defence Intelligence, said in June 2025 that Moscow and Pyongyang had agreed to start preparations for North Korean production of Russia's Shahed-derived Geran and Garpiya drones. Russia has continued upgrading its own drones, including with AI.

The South Korean system's targeting concept is closer to Israel's Harop loitering munition than to Iran's Shahed. Electro-optical and infrared cameras allow an operator to see a target, adjust the flight path in real time and direct the strike. The Shahed, by contrast, generally flies toward preset coordinates after launch.

The ADD localized core technologies such as the data link, while using a British-made engine.

The defense agency used a British commercial low-Earth-orbit satellite communications network during development to exchange commands and data with the drone. The government ultimately plans to build a low-Earth-orbit communications network using domestically developed satellites.

"A Shahed costs about 30 million won ($21,700) and the U.S. Lucas is in the 50 million won range, but a domestically developed loitering munition costing several hundred million won does not fit the cost-effectiveness demanded by modern warfare," Shin Jong-woo, secretary general of the Korea Defense and Security Forum, said.

BY LEE YU-JUNG [[email protected]]