UAS: The Next Big Thing

Never mind NextGen. It's so last year.

UAS

UAS

I see in The New York Times that rugged-individualist Western farmers are using drones to inspect crops and locate lost cattle, defying FAA regulations that forbid commercial use of the ubiquitous gadgets. "Precision agriculture," as this kind of eye-in-the-sky farming has come to be called, seems like a pretty harmless thing — a good thing, actually, if you like food. In ­Japan, thousands of robot helicopters are already being used for crop spraying. However, despite the surging popularity of recreational multicopters, drones have gotten rather bad press in the United States. Their potential for annoying invasions of privacy and for colliding, accidentally or deliberately, with real airplanes has temporarily eclipsed their many beneficial uses.

If the present rules are any indication, I suspect that the greatest challenge facing the FAA in the coming years will not be the implementation of the GPS-based air traffic control system but the progressive integration of pilotless airplanes into it. Technology, especially software design, moves far faster than rulemaking. Schemes of package delivery by drones, of robotic passenger-carrying "personal aerial vehicles" — just take a seat, select a destination and press "start" — or of pilotless cargo liners flying, like geese, in efficiency-enhancing V-formations may not be as fanciful as they seem. For the time being, however, the FAA clamps down on model airplanes.

Professionals in pilotless flight do not use the popular name drone. The ­preferred term is UAS — unmanned aerial system(s) — with RPA, remotely piloted aircraft, often used with specific reference to the military airplanes controlled by faraway pilots who see the view forward and around as though they were sitting in the airplane itself rather than a trailer in Nevada.

Under the current bizarre FAA policy, nonrecreational use of any kind of model aircraft, anywhere out of doors, requires a special authorization or waiver. This absurdity can only have sprung from a frantic impulse to keep things under control while figuring out what to do next. It can't last long. In February, the FAA published a notice of proposed rulemaking with a somewhat more rational, but still patently imperfect, framework for small — under 55 pounds — UAS operations. It requires, among other things, that UAS operate only in day VFR conditions below 400 feet and that they remain within view of their operators — a condition that curtails their most obvious useful function, that of allowing us to see things in places that we cannot reach. I think it will be obsolete before it ever takes effect.

The NPRM assumes that all UAS are radio controlled by a ground- or water-based observer. There are plenty of other possibilities, including fully "autonomous" aircraft. Strictly speaking, autonomous means completely independent of, and impervious to, human control throughout the course of a flight. Realistically, just like present-day cars and manned airplanes, UAS will perform an expanding repertory of tasks autonomously without becoming fully autonomous. Still, the word autonomous occurs only once in the 48-page NPRM, and that is in a chart comparing the proposed rules with current ones in Canada and Australia (places to which, incidentally, Google and Amazon have gone to do their UAS development work unencumbered by U.S. rules). Under the heading of "autonomous operations" Canada and Australia both get a "no" while the NPRM gets a "yes," but ­without further elaboration.

One of the principal areas in which UAS-related research is being done is that of onboard sensors and related software, both for avoiding traffic and obstacles and for recognizing an object of interest like a game poacher or a lost hiker on the ground. Another is software that would allow semiautonomous UAS to adapt to changing circumstances and, particularly, to emergencies, as a human pilot would.

Many people believe that emergencies are one area in which human ­resourcefulness can't be replaced by a computer. Everyone remembers some famous case in which humans saved the day: Dennis Fitch and Sioux City, Chesley Sullenberger and the Hudson. Ella Atkins, associate professor of aerospace engineering at the University of Michigan and herself a private pilot, thinks otherwise. Her research involves pilotless systems and ones on the cusp between pilotless and piloted. Atkins has made the iconoclastic statement that a computer could have landed Sullenberger's Airbus back at LaGuardia rather than in the Hudson River. She intended no disrespect to Sullenberger; it's just that human beings are incapable of the kind of split second analysis of multiple variables that computers are good at. Sullenberger could not gamble; he had to do what he knew was possible.

Experimental software systems have been developed to maneuver a multiengine airplane with power alone if all control surfaces are lost. Others can, after a total power loss, select the most appropriate airport, stretch of straight road or patch of flat terrain and plan and execute a glide to landing while taking into account wind, weather, terrain and conflicting traffic. Such systems, when mature, would always outperform human pilots and would very likely find their first application not in UAS but as backups in piloted aircraft.

Atkins, together with many other professors of aeronautical science, has been a vocal critic of the current FAA rules, which oblige her students to fly their UAS indoors because their activity is not recreational. She believes the FAA's claim of jurisdiction right down to the surface over private property is excessive, and in a formal comment on the NPRM she urged that the legal principle of "immediate reach" be applied to low-altitude UAS operations.

The idea of "immediate reach" comes from a 1946 Supreme Court case, United States v. Causby. Thomas Causby was a chicken farmer whose land lay alongside a military airport. Planes passing low overhead threw his chickens into a frenzied panic. Some were injured, some died, and Mr. Causby's livelihood was endangered. He sued the United States on the grounds that, under ancient common law, if you own the land you own what's under and above it too. The government's use of the airspace above Causby's land therefore constituted an "unlawful taking."

The court agreed with Causby, up to a point. Justice William O. Douglas observed that the old doctrine could not hold in the modern world, for if every landowner owned all the airspace above his property, every airplane would be trespassing much of the time. But Douglas wrote that "if the landowner is to have full enjoyment of the land, he must have exclusive control of the immediate reaches of the enveloping atmosphere." Douglas left future jurists to determine what "immediate reaches" meant, but to farmers wanting to survey their crops and to universities using model airplanes to teach students aeronautics, it would obviously include a few hundred feet above their property. Atkins suggests leaving a layer 300 or 400 feet deep — excluding, of course, places where it would conflict with other flights — in which UAS would be regulated not by the FAA but by ­existing laws regarding other kinds of equipment and conveyances.

Operations close to the surface pose, at present, relatively simple regulatory challenges because there is still little or no air traffic there. But pilotless airplanes are not all 6-pound quadracopters with Hero cams; they can be of any size, with any performance, and will need to operate in all airspace. The task of their designers will be to make them pass a sort of aeronautical Turing test: It should be impossible to tell from their behavior whether they are piloted or not. The task of the FAA, with the help of NASA, will be to reconfigure the National Airspace System so that UAS are no longer square pegs in round holes. It will be the task of the lawyers and insurers to sort out the rest.

Hold on. It may be a rough ride.

Get online content like this delivered straight to your inbox by signing up for our free enewsletter.