The Complex

Semi-autonomous killer drones from around the globe

With the United Nations Human Rights Council debating the development of lethal robots at its meeting in Geneva today, Human Rights Watch is ramping up its campaign to get world governments to preemptively ban the use of killer robots that can decide to attack a target without consulting their human controllers first. Despite the fact that the Pentagon has said that U.S. drones will not be able to fire weapons without "appropriate" levels of human control, HRW worries that battlefield necessity will do away with such voluntary restrictions.

"While a positive step, [DOD's policy on lethal drones] is not a comprehensive or permanent solution to the potential problems posed by fully autonomous systems," reads a May 28 Human Rights Watch statement. "The policy of self-restraint it embraces may also be hard to sustain if other nations begin to deploy fully autonomous weapons systems."

While these real-life Terminators don't exist yet, Killer Apps thought we'd compile a list of robot weapons that are leading the drive toward autonomy. Although the U.S. military isn't looking to field fully self-functioning killer robots, it does want to get its drones to the point where one person can control a fleet of them. Under the current setup, several pilots and sensor operators are needed to fly just one of the military's larger drones, such as the MQ-1 Predator or the MQ-9 Reaper. The same thinking can be applied to ground robots and those that prowl the seas, above and below the waterline.

DreamHammer is a company that makes a system called Ballista, which allows one person to control multiple drones. Its CEO, Nelson Paez, explained to Killer Apps how such a semi-autonomous fleet might be used:

An operator providing medical evacuation support could have a multitude of surveillance assets that finds the wounded and then be able to manage unmanned helicopters and/or ground vehicles going into unsafe areas and quickly picking up the wounded. Another example, would be a single sailor (or a coordinated set of sailors) on a Navy ship on a common control station with the ability to manage unmanned intelligence, surveillance and reconnaissance [ISR] assets that find targets and then utilize strike weapons such as Tomahawk [cruise missiles] to take out a target and use the unmanned ISR assets for battle damage assessment.

So, which unmanned vehicles could actually do such work? Here's a partial list.

First up is the U.S. Navy's favorite drone (and perhaps the sexiest of all drones?), the Northrop Grumman-made X-47B Unmanned Combat Air System demonstrator (UCAS-D). It's the history-making stealthy jet that took off from and performed touch-and-gos on the aircraft carrier USS George H.W. Bush earlier this month. (Not only is the X-47B the first jet-powered drone to launch from an aircraft carrier, it might be the first stealthy jet to take off from a ship via catapult launch.) The X-47B is designed to fly missions semi-autonomously, meaning that a human "mission operator" tells the jet to do things like take off, land, and search for targets without actually controlling the plane as it executes these functions. The Navy wants to use the tech developed for the X-47B to field a class of carrier-borne stealth drones by the early 2020s that can perform long-range surveillance and strike missions with one operator controlling multiple aircraft.

Next is France's nEUROn. Made by Dassault, the nEUROn is basically the land-based, French version of the X-47B. As Dassault says the nEUROn is meant to test the technologies behind flying a stealthy drone capable of autonomously finding an enemy on the ground using its onboard sensors (along with info from other spy assets) and then dropping bombs. nEUROn made its first flight (shown above) at Istres, France on December 1, 2012 and will undergo flight tests there until 2014, when it will move to Sweden for even more tests. Then it will move to Italy, where it will be put through yet more tests designed to see how stealthy it is and how good it is at dropping bombs.

Both the UK and China are getting ready to flight test their own stealthy UAVs soon. While the British jet, the BAE-made Taranis (shown at the very top of this article), is designed to be semi-autonomous, it's not clear how much, if any, autonomy the Chinese jet is designed to have.

Although the United States is not developing fully autonomous killer vehicles, Israel already has a semi-autonomous one in operation; its scary little killer robot car, the Guardium. As Killer Apps showed you last November, the Guardium is patrolling the border between Gaza and Israel. It uses a variety of sensors to navigate and looks for suspicious activity, and it can even "quickly respond and hold the suspicious elements back until manned troops arrive, or use various forceful methods to eliminate the threat," according to the Israel Defense Forces' blog. The best part? The Guardium can "run patrol on predetermined routes without human intervention." Should it find something it doesn't like, it can "react to unscheduled events, in line with a set of guidelines specifically programmed for the site characteristics and security routines," according to its manufacturer, Israel-based G-NIUS Unmanned Ground Systems. 

One decades-old weapon that you might not think of as a killer robot -- but that the UN special rapporteur on extrajudicial summary or arbitrary executions considers a potential "lethal autonomous robot" -- is the Phalanx Gatling gun, used by the navies of the U.S. and its allies. The gun is known by its acronym CIWS (pronounced "see-whiz" for Close In Weapons System), and it has a land-based counterpart, the C-RAM (Counter-Rocket, Artillery and Mortars). CIWS is a ship's last-ditch defense against air attack, using a built-in radar to automatically track and fire on incoming planes or missiles. C-RAM uses similar technology to try to shoot down rockets, artillery, or mortars that are raining down on U.S. bases. While CIWS isn't as glamorous as moving robots, there have been several incidents of the robo-Gatling guns accidentally firing on U.S. Navy ships and aircraft. (Click here to read about them.)

BAE Systems, U.S. Navy, IDF, Wikimedia Commons, Dassault Aviation

National Security

DOD says don't worry about hackers accessing key U.S. weapons designs

The Pentagon is denying that any real damage resulted from hackers accessing the designs for more than 24 major U.S. weapons systems.

"We maintain full confidence in our weapons platforms," reads a just-released statement from DOD Press Secretary George Little. "The Department of Defense takes the threat of cyber espionage and cyber security very seriously, which is why we have taken a number of steps to increase funding to strengthen our capabilities, harden our networks, and work with the defense industrial base to achieve greater visibility into the threats our industrial partners are facing. Suggestions that cyber intrusions have somehow led to the erosion of our capabilities or technological edge are incorrect."

That's right, DOD claims that all is well despite the fact that, according to a classified version of a Defense Science Board report, hackers have accessed designs for dozens of weapons systems, ranging from the F-35 and F-22 stealth fighters to numerous air-defense missiles, advanced communications technologies, lasers, RC-135 Rivet Joint spy planes, and even the Navy's Aegis anti-missile system.

Little's statement comes a little more than two months after the U.S. Intelligence Community listed "Cyber" as the top security challenge in its annual Worldwide Threat Assessment, saying that U.S. adversaries are "almost certainly" using cyber espionage to catch up to the U.S. military:

Highly networked business practices and information technology are providing opportunities for foreign intelligence and security services, trusted insiders, hackers, and others to target and collect sensitive US national security and economic data. This is almost certainly allowing our adversaries to close the technological gap between our respective militaries, slowly neutralizing one of our key advantages in the international arena.

His comments also come after years of warnings by U.S. government officials -- from U.S. Cyber Command chief Gen. Keith Alexander to House Intelligence Committee Chairman Mike Rogers -- about the damage caused by cyber espionage and crime.

We've written plenty on allegations that Chinese hackers have stolen the plans for various U.S. weapons systems and have pointed out that China's stealth jets bear a suspicious resemblance to U.S. stealth planes like the F-22 and F-35. Designs for the F-35 were reportedly hacked by Chinese spies in an incident that may have contributed to the redesign of the jet's computerized maintenance system.

Perhaps Little's message is simply a display of false confidence, or perhaps the U.S. has made enough changes to programs accessed by hackers that it's not worried, or maybe it simply fed them the wrong information.

Some cybersecurity experts are already calling on U.S. firms to start making it costly for hackers to steal information from them by poisoning the virtual well.

"We have to get the Chinese and the other adversaries off the idea that when they exfiltrate the data out [of U.S. networks], that it's pure," said James Mulvenon, vice president of intelligence at security consulting firm Defense Group, Inc., during a speech last week at the Center for Strategic and International Studies. "They believe this is Ultra, this is the most profoundly successful intelligence operation they've ever had. They believed with metaphysical certainty, up until recent times, that what they're exfiltrating is actually true."

Mulvenon seems to be suggesting that the U.S. is already pumping false info to cyberspies.

"Using deception and poisoning the well and things like that in terms of the data exfiltration is obviously not new. It can be technically difficult, but we've seen [that] the tried and true methods we've used in the counterintelligence and counterespionage realms have really helped us," Mulvenon added.

At the same event, Shawn Henry, a former FBI cybercrime investigator who heads the services division at cybersecurity firm Crowdstrike, said that American businesses need to start "being proactive...being able to raise the cost to the adversary. Right now there's no cost, the risk is about zero because people have been called out on it for years and years and nothing is happening."

"Denial and deception is key -- changing the way we look at these things, being proactive on the network, not in an offensive, aggressive way" but by creating capabilities that "make things more difficult for the adversaries" by giving them bad information and quickly identifying the attackers, he said.

Still, this approach might backfire if not executed by anyone but the most sophisticated cyber security teams warns Dave Merkel, chief technology officer of cybersecurity firm, Mandiant.

"I myself am skeptical of those approaches, when I go take a look at a large organization and the challenges it has managing its own legitimate information, and then you talk about managing legitimate disinformation and being able to tell one from the other and being able to make decisions based on what happens with it seems pretty far fetched," Merkel told Killer Apps. "Those kinds of techniques can be effective in highly-targeted ways, used by specialists to get some particular result like learning more information about an adversary . . . but as some kind of broad-based defense or mechanism to change the economics of stealing digital information, I just don't see it."

Mandiant famously published a report in February detailing the exploits of an alleged Chinese-military hacking group against U.S. businesses. Merkel said the latest news about DOD weapons designs being hacked is nothing new.

"This just verifies what we're seeing within our own client base," said Merkel. "I wasn't even mildly shocked."

Wikimedia Commons