Eric Adams WIRED
America’s Plan to Somehow Make Drones Not Ruin The Skies
May 3, 2017
  • Share
  • IF YOU CAN’T wait for the day drones plop packages on your porch or a flying car whisks you to work, you should know that the hold-up isn’t technological, but technocratic. Before these future flyers can take off, they must learn to play by the rules of the sky.

    That means communicating with air traffic control and other aircraft, spotting and avoiding threats, and generally knowing what to do when things go sideways. Making all of this happen demands whole new levels of capability—not just from the aircraft, but from the sprawling system that oversees them.

    The good news is, change is coming. Even though commercial drone technology barely existed a decade ago, regulators are hustling to integrate it into the national airspace. The FAA tapped Intel CEO Brian Krzanich to lead its Drone Advisory Committee, and established seven test sites to explore drone flight management. NASA is supporting the quest with its unmanned aircraft traffic management research program.

    Although everyone is moving fast, they’re shunning the “break things” part of the of the Silicon Valley ethos. Safety trumps efficiency and technological advances. “Our challenge is to find the right balance where safety and innovation co-exist on relatively equal planes,” FAA chief Michael Huerta told the Unmanned Aircraft Systems Symposium in March. “As we move toward fully integrating unmanned aircraft into our airspace, the questions we need to answer are only getting more complicated.”

    The Current Regime

    So far, the primary answer has been, no fun for you. Anyone operating a drone for commercial purposes must first pass a test covering traditional pilot know-how like wing load factors and airspace regulations. Drone pilots must keep their flyer below 400 feet, away from crowds and airports, and within their line of sight—restrictions designed to keep drones out of trouble. The FAA occasionally waives some restrictions with an eye toward seeing which rules it could loosen, but some people say overzealous regulation will keep the tech from reaching its potential.

    “There are so many applications that will benefit from drone use,” says Mark Barker, director of business development and marketing at the Nevada Institute for Autonomous Systems, which the state government created to expand the use of unmanned air systems. “Energy companies can inspect 50 miles of power lines with winged robots instead of human-piloted aircraft. Drones can conduct search-and-rescue, fly filming missions, survey environmental hazards, and act as couriers. They’re more efficient and more cost-effective than human-piloted aircraft.”

    The FAA sees good reason to move cautiously, given basic questions remain unanswered. How will the aircraft respond when it runs out of power or experiences a failure? How will it communicate with air traffic control to approve a flight path or line up for a landing? And how will it avoid smacking into other aircraft?

    “If there’s no pilot in the aircraft scanning the horizon, and it’s not big enough for its own radar, how will they detect other aircraft?” says Ed Waggoner, director of NASA’s Integrated Aviation Systems Program. “How do you get information to the unmanned aircraft or the ground-based pilots-in-command in remote-flying situations to avoid other aircraft that may not be clearly announcing their intentions?”

    Moving Forward

    New, miniaturized sensor technology will be key here, particularly the ADS-B positioning systems increasingly common in commercial and civilian aircraft. That technology is approaching approaching drone-friendly dimensions and power requirements, but other problems remain. “In the case of what we call ‘non-cooperative’ aircraft—those flying without those systems either deliberately or because they’re not required to—we need to determine whether ground sensors or airborne sensors in other aircraft represent the best strategy for tracking them,” Waggoner says.

    Communicating with air traffic control presents its own problems. The FAA must determine how onboard flight management systems will function, and how those data-based, nonverbal systems will “talk” to controllers on the ground. It could be that the systems operate in a hybrid fashion (with human aircraft operators on the ground assisting otherwise autonomous aircraft through certain phases of their missions) or the successful integration might be contingent on the arrival of the next generation of air traffic control technology.

    That happens to be in the pipeline, as well, via the FAA’s ongoing NextGen modernization program. This concurrent evolution is both a complication and a blessing for planners working to merge unmanned aircraft into the national air system. “We’re working with the FAA to ensure that our own work with UAS integration is forward-looking, so we’ll be ready for how air traffic control will function in the transportation system a decade or more from now,” Waggoner says. “But we also need to be safely testing these systems in a realistic, current environment.”

    That’s where the FAA’s seven test sites come in. For instance, later this month, Nevada’s Institute for Autonomous Systems—the largest of the seven, with the ability to operate state-wide—will participate in a “Technical Capability Level” evaluation at its Reno facility. It will test a variety of traffic management systems on fixed-wing airplanes and multirotor copters, with flights up to 1,200 feet altitude and across several miles of approved airspace. (Other test sites around the country will be similarly engaged, running their own tests with various commercial and research partners.)

    In Nevada, researchers will conduct long-distance aerial survey, package delivery, and emergency response missions. They’ll also try out ground-based sense-and-avoid systems. In January, they’ll throw in “non-cooperative” aircraft—those not identifying themselves—over moderately populated areas. Sometime after that, they’ll expand the test to denser urban areas and explore tasks such as news gathering and package delivery.

    As they move toward full freedom, these aircraft must rival humans not just in skill, but in judgment, knowing what to do with the information they collect. “It’s not just sensing and knowing, but the vehicle needs to know the rules-of-the-road when flying—go left if you’re on a collision track, go over there to avoid contact, plan your path this way, etc.,” says Richard Pat Anderson, director of the Flight Research Center at Embry-Riddle Aeronautical University. “That requirement is not baked into the current FAA vision for NextGen, so it will need to be there, as well.”

    When the FAA has finally reached a solution that answers all these concerns—there’s no specified timeline—it will gradually alter the regulations to permit unmanned and autonomous aircraft to zip around the skies, hopefully without causing trouble. Only then will you be able to whip out your smartphone and order some sneakers or whistle up an autonomous electric air taxi to take you to the mall to buy them yourself.