Autonomous Underwater Robots Could Cause Deadly Accidents, Warns UN Think Tank

The “U” in U-boat now stands for “unmanned”

The next sea change in naval warfare may be unmanned, autonomous robots. In a new white paper published by the United Nations Institute for Disarmament Research, an internal think tank, the fear isn’t the reaper, but the self-steering submarine. The UN’s biggest concern: not only are conflicts likely to happen in coastal waters or contested seas, but the unique nature of the ocean means that it’s well-suited for war with autonomous killing machines.

One reason the sea is ripe for robot warriors is because remote control under water is harder than it is in the air. While current drones navigate by GPS and relay signals through radio or satellite uplink, those signals don’t travel as well through the sea. An autonomous submarine, instead, could operate for days underwater, returning to the surface to transmit the data it collected and receive new instructions. Boeing has already built and the U.S. Navy is looking to buy a similar, maybe identical, large autonomous submarine to scout the sea and report back. Rather than competing for bandwidth in a crowded sky, unmanned sea robots will work on their own, without a need for step-by-step handholding.

Titled “The Weaponization of Increasingly Autonomous Technologies in the Maritime Environment: Testing the Waters,” the white paper isn’t just concerned with submarines. Aegis autonomous anti-missile weapons have served on American ships for decades. These missile systems work faster than humans possibly can to detect and fire upon incoming targets. However, their risk became clear in 1988 when an American ship accidentally shot down an Iranian airliner over the Persian Gulf, killing all 290 people on board.

Deadly underwater autonomous weapons date back farther than that. Sea mines, explosives placed in the ocean and designed to destroy ships, date back centuries, and since World War II, they’ve damaged four times as many U.S. Navy vessels as anything else. In a twist on the sea mine of old, DARPA wants underwater pods that wait until activated, and then launch drones from the ocean.

There are several big risks that come with newer, smarter, and more sea weapons. The first is simply tracking them; sea mines from WWII remain a problem in parts of the ocean to this day, and staying current with every deadly weapon, especially the ones a nation might want to keep secret, is a challenging task. Another risk is accidents, like the 1988 misfiring of the Aegis system. More so than any other weapon, an autonomous machine risks acting in a way that the humans responsible for it don’t want. And there are environmental concerns. We already know bears don’t like drones flying near them; how well will whales take to mechanical beasts moving in their midst? Or what if opposing robots fighting a human war under the sea poison a protected reef in the process?

This report joins others, like one from Washington DC’s Center for New American Security, that seek to define the terms of autonomous systems before they’re widely in use. Without even a basic attempt by policy makers to understand the issue, the future is likely one of robot ships, borne into war on tides of darkness.

[Via Defense One]

Kelsey D. Atherton

Kelsey D. Athertonis a defense technology journalist based in Albuquerque, New Mexico. His work on drones, lethal AI, and nuclear weapons has appeared in Slate, The New York Times, Foreign Policy, and elsewhere.