As the American-led wars in the Middle East wind down, the troops are being called home and they’ve been joined on the return trip by another element associated with the military occupation: drones. Thanks to legal reforms shepherded through the US Congress last year (with a push from the defense sector) the Federal Aviation Administration (FAA) is set to begin licensing drones for commercial use by 2015, and the US may have as many as 30 000 drones in the domestic skies by 2020. These developments will have far-reaching effects across many areas of our daily lives and raise questions of privacy and public safety. With an industry already measured in the billions of dollars, there’s mounting pressure to build up the drone economy without spending time answering difficult questions. But the answers to these questions are important. Are we ready for the coming drone revolution?
After all, even the term ‘drone’ itself is a bit murky. The military distances itself from the word—and its implication of unskilled and mindless pilots—and prefers the more unwieldy ‘unmanned aerial vehicle’ (UAV) or ‘remotely piloted vehicle’ (RPV). Congress and the FAA go with ‘unmanned aircraft,’ or refer to a full ‘unmanned aircraft system,’ to describe the vehicle, the ground control unit, and any other supporting devices. Purists insist that it’s not a drone if it can’t fly autonomously, ruling out most model airplanes flown by hobbyists.
However, for the purposes of a discussion about the impact drones will have on our daily lives, it makes sense to consider not just autonomous and aerial vehicles, but really any unmanned and mobile robotic machines. This broad definition includes the military drones we read about in the news, like the Predators and Reapers conducting surveillance and strikes overseas, but also much smaller commercially available hobby drones, Google’s self-driving cars and even the humble Roomba robotic vacuum.
Casting such a wide net is useful, because our considerations need to cover a lot of territory. The drone revolution is about more than aerial autonomy, or removing the human element from the cockpit. Rather, it’s about personal robotics increasingly becoming a part of everyday life. The defense sector has a head start on constructing aircraft with sophisticated autonomy software, but as the world of drones opens up to more commercial activity, the market will produce things we never expected.
In that sense, drones symbolize the world of personal robotics, poised right on the edge of explosive development. Chris Anderson, the former editor of Wired Magazine, made a splash in the media world when he left that job to work fulltime on a business catering to drone hobbyists. In explaining his move, he’s joined a growing group of people drawing a comparison to the personal computer revolution that took place in the decades after the Homebrew Computer Club first started building kits in garages in Silicon Valley. In other words, drones and the technologies they represent may end up being the Next Big Thing on that scale—and as in a previous era, could make today’s entrepreneurs very rich for their efforts.
It might be tempting though to dismiss personal robotics as a valuable niche without much effect on regular individuals, say, or as useful for the military but irrelevant to civilian life. But it bears reflecting that even the pioneers of the early computer famously underestimated their eventual effect on society. There’s a remark often attributed to Thomas Watson, the CEO of IBM Computers through much of the first half of the 20th century: “I think there is a world market for maybe five computers.” Even those in the know are unlikely to foresee the innovations and developments that can happen as the cost of robotics components come down and people start experimenting.
If personal robotics live up to that promise, the results are bound to be mixed. Intertwined with hopes for the future of drones are concerns about their effect on personal privacy, about their operational safety and their vulnerability to exploits by hackers.
Potential Privacy Problems
The privacy problems posed by drones are as apparent as they are thorny. Whether in the hands of average citizens, police officers, or the federal government, the ability to cheaply and anonymously conduct aerial surveillance challenges most people’s conception of personal privacy. The rise of drone technology is matched, too, with a corresponding improvement in cameras and image analysis, so even as the vehicles get smaller and faster, their internals will be getting smarter, capturing more detail and running software that can recognize faces or even evaluate situations. Already, there are drone systems running in the Middle East that, from tens of thousands of feet in the air, can record entire towns, or read the text off a milk carton lying on the ground.
These facts tend to evoke such visceral privacy goosebumps that some scholars—most prominently the University of Washington’s Ryan Calo—have argued that drones will end up being a privacy catalyst. That is, people will push for privacy laws stronger than those on the books in order to address the current threat. Since Calo made that prediction nearly two years ago, it has already begun to bear out; on the federal level, state level and even in many cities, legislators are making proposals that push for additional privacy protections.
Such proposed laws address a possible incompatibility between drones and the Constitution’s Fourth Amendment, which prohibits unreasonable searches and seizures. As in other areas touched by technology, the concept of a ‘reasonable expectation of privacy’ is coming under closer consideration. At one time, ‘reasonable’ was bound as much by physics as by law: there’s not yet an explicit warrant requirement for constant aerial surveillance of a person, but it would have been impossible or prohibitively expensive to conduct before. Within a few years, it may be trivial.
The Supreme Court recently addressed this same interplay between laws of physics and laws in statutes in a landmark privacy case called United States vs. Jones. The case addressed evidence gathered by attaching a GPS tracker to the suspect’s car, collecting location data continuously for a month. The government argued that such a practice was not functionally different from tailing the same suspect with agents, which does not require a warrant. The Court disagreed. There is such a substantial difference in cost, ease of use, and even capabilities that this difference is one of kind, not just of degree. (When that decision came out, the FBI was forced to turn off some 3 000 GPS trackers it had in the field).
That’s a useful case to have in mind when police leadership or politicians argue that the issues raised by drones are no different from those raised by helicopters or small planes. Neither of those vehicles requires a warrant to use, either.
Even if these privacy issues get resolved, there are still concerns about the operational safety of these machines. Despite the long way that drone technology has come in the past decade, there are still major reliability problems. These problems will have to be addressed before they can be deployed over crowds of people, or even really above private property.
Not only are the ‘sense-and-avoid’ systems designed to prevent mid-air crashes still relatively primitive, but airspace is likely to get significantly more crowded. Without a human pilot in the loop, collisions are all but inevitable. There have already been some spectacular crashes, mostly in Afghanistan where the larger drones are deployed. In Somalia, too, a recent United Nations report documented some crashes and narrowly averted disasters. And of course, that problem is also coming home. Last summer in Maryland, a Navy drone (with a price tag of $176m [USD]) crashed into a river normally open to recreational boat traffic.
In addition to operational safety considerations, drones may become a proxy for long-standing questions of police force. An enthusiastic sheriff in Texas has become one of the most vocal proponents in the US for arming police drones with ‘less lethal’ charges like bean bags or tasers. Of course, the trend of militarizing police is much broader than its applications to personal robotics, but already activists see that battle as the next major front.
It should be no surprise, then, that in cities like Oakland and Seattle, the sites of gruesome scenes of police brutality in 2012’s Occupy demonstrations and 1990’s ‘anti-globalization’ riots, opposition to police drones is most fierce. In both cities, grassroots citizen groups have developed to demand more transparency about the plans for drones, and in some cases to seek a moratorium on drone use altogether.
Many of the safety concerns that drone opponents identify are more pronounced with the larger military drones deployed overseas than with the relatively tiny (and unarmed) vehicles currently used in the US. But it’s a mistake to write off these problems as something for the military alone to worry about. Already the Department of Homeland Security (DHS) has some 10 Reaper drones operating as border patrol. More alarmingly, DHS has loaned these drones to local and state police departments on a handful of occasions, under terms that are currently not available to the public. The Electronic Frontier Foundation, a San Francisco-based organization working to protect fundamental rights regardless of technology, filed a lawsuit for that information, but the point stands—the safety issues raised by drones overseas are increasingly the ones we have to worry about here at home.
Vulnerabilities and Exploits
All of the problems outlined so far presume a standard relationship between operator and drone. But one more complicating factor is that these machines, in some cases powerful networked general-purpose computers with wings, are themselves vulnerable to a host of exploits. Unlike more traditional hacking scenarios, the consequences of a drone being compromised can be both digital and physical.
Most dramatically, drone navigation units are vulnerable to two different kinds of attacks on their GPS systems. ‘Spoofing’ entails sending strong (but faked) GPS signals towards a drone, so that it starts following your lead instead of its programmed directions. It wasn’t until last year that a successful spoofing attack was conducted against a drone, and the researcher who managed it—at a DHS facility, in front of an audience of impressed agents—called it very sophisticated. But still, he achieved it with just $1000 (USD) worth of equipment,and the stakes get higher each day.
For now, military GPS uses encryption that renders it invulnerable to any known spoofing attack, but even it is susceptible to ‘jamming.’ In a jamming attack, the drone is overwhelmed with signals to the GPS antenna. The encryption ensures no fake signal is taken as the true one, but the true signal can’t get through either. When a US drone went down over Iran in late 2011, an official statement from the Iranian government indicated that its military had jammed the drone’s GPS signal, but that has not been confirmed.
Aside from attacks on the navigation of the drone itself, there is the security of the payload it’s carrying and the network that it transmits on. Outside of the military, in a private context, this information may be even more valuable to attackers. The obvious targets here are the cameras or sensors equipped on a drone. These attacks aren’t hypothetical, either. An investigation prompted by a handful of documented cases of militants in Iraq with apparently captured videos on their laptops revealed that a piece of $26 (USD) off-the-shelf software was capable of intercepting feeds from US military drones. If the military isn’t capable of securing its videos against leakage like that, why should we expect police forces or private companies to be? And again, this problem only gets more complicated as more and more sensitive information is being recorded.
Towards a Better Future With Drones
After looking at all of these real and serious problems with domestic drones, it may seem like the future is grim—either for the drones, or for us. But in truth, the possible benefits from drones are equally numerous. That’s why it behooves us to work out solutions.
Citizen journalism is one area that, as with previous advancements in communication networks and recording equipment, stands to benefit greatly from a more widespread availability of drones. Already, innovating journalists are taking action, in some cases building their own models in order to capture a story. In early 2012, one drone hobbyist outside of Dallas used a point-and-shoot camera and a $75 (USD) airframe to document the gruesome evidence of an improperly operated meatpacking plant. Even if that footage had been possible to uncover without a UAV, it would have been prohibitively expensive for an amateur. Amidst the constant surveillance of Occupy Wall Street, one citizen journalist developed plans for an ‘Occucopter,’ a drone that could record the police and broadcast footage real-time on the web. These sorts of concepts are hitting the mainstream too, getting academic recognition, for example, at the first ever Drone Journalism Lab at the University of Nebraska-Lincoln.
When personal technologies are on the brink of real widespread popularity, they face two paths forward: they can either entrench existing power dynamics, or they can work to reverse them. It can take a long time before which path is which becomes clear, and many technologies do some of both.
But people like Anderson hope for the latter. Like the Internet, which has already managed to connect people and lead to real improvements in many people’s lives, drones and personal robotics could work to empower us. Of course, that’s not without risks, and not even the most benevolent technology has been purely a force for good. But despite the dangers, drones may yet be a net positive in society.
In the case of the Internet, which brought personal computing to the next level, it was a combination of thoughtful (and political) engineering and a democratic governing body that allowed the technology to develop in the right way. If we want drones to develop in the same direction domestically, we need to continue to push for thoughtful development of the technology, and we need to consider whether regulation on the local level would be effective to step in where technological limitations have been brushed aside. If we do that correctly, the future of drones and personal robotics may surprise us with its novelty, but still meet our expectations for the society we want to live in.