Search and Rescue and the Spread of UAVs

Sorry for the light posting of late – the Anderson family is currently in the Sierra Nevada, the eastern side out of Bishop, California, on God’s own highway, the Empty Quarter of Highway 395, which runs north-south from southern California all the way up the eastern Sierra and beyond.  It is both the most beautiful and most varied countryside you can imagine.  If the gods loved you, you would be here, as are we.  Unfortunately not spending enough time, however, so we are just doing various day hikes.

There is not a lot of international law in the eastern Sierra Nevada.  There is an important body of sovereign nation law, given that there are several Indian tribes and tribal lands up and down the Owens Valley, including the Paiute-Shoshone tribal lands in the center of Bishop.  But one feels somewhat removed from the Law of Nations.  However, I thought I would share one conversation with one of the rangers here in the national park.  She remarked that the ranger services – national parks, national forest, etc. – had been watching with great interest the growth of unmanned aerial vehicles (UAVs) in civilian use.  So far this includes things like crop dusting and surveillance.  Because the air bases that command some of the UAVs are located in Nevada, relatively nearby (in the Empty Quarter, that might mean 300 miles of desert driving, of course), there is a lot of awareness of UAVs and their potential – University of Nevada, Las Vegas has just begun a program to graduate UAV controllers, for example.

When people talk about surveillance UAVs, they are typically thinking about border patrol, but here, the park services are thinking about fire patrols – an immensely expensive task from aircraft now, because of the vast areas to be surveyed in real time – but worth it because the faster the fire is spotted, the better the chance of containing it before it spreads.  LIkewise, search and rescue for lost and injured back country hikers.  That one is somewhat ahead of existing technology, for what the park services would ideally like, because flying in the steep valleys and canyons is difficult and hazardous now, but UAV technology is not sufficiently up to speed to take over those tasks.  But it will happen soon, as smaller UAVs that are more like large birds can be deployed in difficult, deep, or narrow spaces.  Likewise, as the sensor technology gets better, cheaper, and more available, it will be easier to find a single lost hiker using not just things like infrared signatures, but sensor arrays that are … well, if they exist, they are still only available to the military.

Point being – this will not be a surprise to OJ readers who understand that this site is UAV-targeted killing central information station – that UAVs are going to spread rapidly and widely across a huge array of tasks and functions currently carried out by manned aircraft.  It will happen because UAVs will be so much cheaper, efficient, and in many functional aspects superior to using people in airplanes.  The impetus will rapidly turn from being military, as it still is now, to civilian.  Everybody, everywhere in the world will shift that direction.

I raise this because there is a meme that still circulates with some velocity in the international law community, journalists, and others, that the US is risking setting off some kind of UAV arms race by its increasing roboticization of conflict – not just UAVs, but ground vehicles, and so on.  I don’t think that’s right; the meme fundamentally misunderstands the technology and its application.  Rather, UAVs are going to spread across a very wide range of aviation in any case, in which military uses will just be one of them.  The same technology, cost, safety, efficiency, and so on, drivers that push for fire surveillance in the Sierra Nevada will be exactly the same ones that drive the military to use the technology.  One can call it an arms race, I suppose, but only if one imagines that it is all about military use, otherwise it is a misleading way of thinking about the technology.

A better way to think about this is to go back to what make robots robots.  In general, there are three conceptual pieces:  A locomotion function or means of gross movement or action in the world; computing and central processing power to be able to analyze; and sensors to bring in streams of data which, being analyzed, result in some form of gross mechanical action.  (In the case of US military UAVs, we can add an additional piece that brings them into an intersection loosely with ‘cyber’ – the communications net that allows them to be piloted over Afghanistan from the US.)  Focusing on the UAV’s gross locomotion part, the flying part, and saying that it will lead to an arms race in which everyone will want one and arm it with a missile misses the point.  There is no arms race about that – the technology for flying remotely has been around for decades; anyone who wants to build one can do so at a hobby shop.  Putting a missile on it is child’s play, literally – presumably no one would be so politically incorrect as to propose building a Predator with a missile as the next high school robotics competition for high school teams, but apart from political sensitivities aside, one reason is that it’s just too darn easy.  Flying is easy; making a machine that walks up stairs is hard.

Everyone will have UAVs because everyone will want them for so many, many things, mostly unrelated to military or police missions.  Any government that wants to arm one with a missile will have no difficulty doing so.  The real technology issues are not with flying, or with weaponization – or even with computing power.  That’s all off the hobby kit shelf.  No, the real technology issues arise with sensors.  One robotics scientist in Silicon Valley told me last year that it was largely unrecognized, but the real advances in technology of the past decade had not been in computers as such, but in sensors and controllers, ranging from new ways and kinds of bringing data streams online to direct neurological, direct brain control of robotic limbs for amputees, and so on.

But now, note the issue.  Some of this technology is classified for military R&D; other parts are not.  The importance of robots outside of the UAV context are immense in large part because the Baby Boom generation does not have sufficient children to see us off to our reward; we are going to slide into dementia and be cared for and comforted by cuddly robotic dolls that we will think are human, to judge by where things are going in Japan.  In the US, we are not so aware of this, yet, although it is striking that the Times and the WSJ have both moved on in their robotics coverage from targeted killing via UAVs to much more friendly news stories about Alzheimer’s patients in Japan being soothed by robot plush dolphins.  Dolphins that will be smart enough to monitor medical conditions and call 911 if needed, to take obvious examples, or monitor whether a patient has taken the meds, or any number of things.  What lies behind this is sensor technologies.

In an armed conflict context, however, it is questionable how many of the fighting forces in the world, state or non-state, will feel any great obligation to minimize collateral damage or attempt to more and more affirmatively id a target before striking.  If you don’t feel that obligation – I would estimate that the countries involved will be the US and Israel, and the rest of NATO only insofar as it ever intends to do any more fighting, but in any case, they will simply acquire US technology.  China will likely do so, because it would at least want every capability, and because it can most likely steal the technology and reverse engineer any missing parts.  But either sensor technology will spread across civilian uses, such as elder care robots, so as to make the concept of an ‘arms race’ moot, or else the number of countries that will be “racing” to have such technologies will be almost entirely limited to countries that (a) fight and (b) care about the rules.  That makes the list frankly pretty short.  It is possible that India might join that list, along with Taiwan, South Korea, and a handful of others in Asia.  But there will not be an “arms race” around sensors, because they are useful primarily for reasons related to more discriminating targeting, and the militaries in the world interested in that is not a long list.

Will there be an evolution of arms around UAVs, then?  Yes, but not likely along those parameters.  The likely arms race is along a quite different one.  Predators are slow and noisy for targeted killing; it will not take long before some party – Iran – begins doing what the US did via the CIA in Afghanistan against the Soviets, and supplies rudimentary surface to air missiles to attack the drones.  The arms race will get underway in the classic evolution of protecting air dominance.  The Predator, for example, might launch not a missile, but instead a still smaller drone with a single-person weapon, specifically designed for up close use.  That will be a function not of flying technology or weapons technology, however, but, once again, sensors.  But an arms race over air superiority is not one that has the implications for the supposed dangerous spread of this new military technology – introducing dangerous new dynamics between India and Pakistan, for example – that numbers of commentators seem (still) to imagine.

I am returning to the solitude and off-lineness of the mountains.