Amazon, Uber and other tech giants want to fill the skies with small autonomous aircraft ferrying packages and people from place to place. For that to happen, these robotic drones—also called unmanned aircraft systems (UASs)—need an air traffic control system to keep them from crashing into buildings, human-piloted aircraft or one another. NASA is developing a UAS Traffic Management (UTM) network with several other organizations that the group plans to finish testing next year. Uber, in particular, has a lot riding on the UTM’s success—the ride-sharing company made several announcements last week to promote its proposed uberAIR taxi service. Big questions remain, however, as to whether and when any monitoring and management system will be able to handle the expected volume of large self-flying aircraft, which will be traveling great distances to deliver everything from pizzas to passengers.

Uber is onboard with NASA, at least. The company announced at its Elevate aviation conference in Los Angeles on May 8 and 9 it had signed an agreement to provide NASA with details and data about the inaugural uberAIR service it has planned for Dallas–Fort Worth. In return, the agency will use Uber’s data to make computer simulations of small passenger-carrying aircraft flying over the Texas Metroplex during peak air traffic times. Uber will analyze those simulations to help plan air taxi management in the already crowded skies over Dallas as well as Los Angeles and Dubai—the other cities hoping to start testing uberAIR by 2020.

Uber is targeting urban areas that have a population of more than two million people and a density of more than “2,000 people per square mile,” according to documents on Uber’s Web site. The cities must also have a large and dispersed layout that allows air taxis “to offer significant time-saving benefits at speeds of” 240 to 320 kilometers per hour. The company also points out flights will go from “node to node rather than point to point,” meaning there will be specific—rather than random—pickup and drop-off sites.

Uber also unveiled design specifications at its Elevate conference for the electric aircraft that will serve as the workhorses for uberAIR, which the company hopes to fully launch by 2023. Uber wants vertical takeoff and landing vehicles that can fly up to about 320 kph at a cruising altitude of about 300 meters between rooftop “skyports,” dedicated platforms where flying taxis can take off or land as close to its customer’s final destination as possible (pdf). The vehicles will need to travel up to 100 kilometers on a single battery charge and come equipped with four sets of electric-powered propellers dedicated solely to takeoff and landing. A fifth propeller—on the tail—will provide thrust for forward motion once the craft is airborne. The company plans to fly the taxis autonomously at some point, but the service will initially use human pilots.

Head in the Clouds?

NASA began work on the UTM in 2015 to identify technologies and procedures to help drones fly safely at altitudes up to 120 meters—airspace not typically monitored by the Federal Aviation Administration (FAA). The four-stage UTM plan commenced with demonstration flights over a rural area—where the UASs were not permitted to fly beyond where their pilots could see them—and has progressed to more ambitious tests that extend the distance between pilot and aircraft. NASA is investigating airspace design (such as creating routes or lanes) and geofencing software that uses GPS or radio signals to prevent drones from flying over certain areas. This year NASA is testing technologies that maintain safe spacing between these aircraft over moderately populated regions. The final phase next year will focus on UAS operations in higher-density urban areas for tasks such as news-gathering and package delivery; the FAA does not currently permit drones to be flown over crowds. Once testing is complete, NASA will hand over the UTM to the FAA to implement alongside existing air traffic control for human-piloted aircraft.

Automated UAS air traffic control has a long way to go before it is ready for everyday use, says David Ison, an associate professor at Embry–Riddle Aeronautical University’s Worldwide College of Aeronautics. Some of the sensing and communications technology will be installed on the ground whereas other devices will be mounted on the aircraft themselves to broadcast coordinates, images, altitude range and other data that help drones to maintain safe distances from their surroundings. “The problem is, if we have lots of drones in the air at once, how do we keep these systems from being overwhelmed?” Ison says.

Even after the UTM is fully tested it will likely be years before the FAA can implement a system that scales quickly enough to accommodate the expected demand in commercial drone delivery services across the U.S., says Parimal Kopardekar, NASA's senior technologist for Air Transportation Systems and principal investigator for the UTM project. There are currently up to 6,000 drones flying through U.S. airspace at any given time but that number will likely increase 100-fold in some places once the FAA opens the skies to drone-based commerce, Kopardekar says. In a sign of things to come the U.S. Transportation Department last week announced 10 sites—including Alaska, North Carolina and Oklahoma—where it will allow a greater range of tests as part of its drone integration program than are generally permitted by federal aviation regulators, including flying drones at night, above people and beyond an operator’s line of sight.

Accommodating large numbers of delivery drones buzzing dozens of meters in the air will be difficult enough, but self-flying taxis pose several “interesting challenges,” Kopardekar says. Those include certifying air taxis for safety, making sure noise from those vehicles does not grate on the communities they serve, installing cybersecurity measures to protect against hackers and integrating commuter traffic with existing plane and helicopter operations. There are already 10 times more drones than manned aircraft registered to fly in U.S. airspace, and logistics become more complicated over urban areas. “We learned through flight tests, if you are operating where the winds or updrafts are really heavy, your vehicle can really bounce around—as much as [30 meters],” Kopardekar says. “We also learned about the impact of high altitude on performance because the air is thin and cold, which means the battery won’t last as long” as it does closer to the ground. In addition to those considerations operators must have a plan for landing safely in crowded areas during a malfunction or emergency, he adds.

The FAA will roll out the UTM in stages, beginning with the Low Altitude Authorization and Notification Capability (LAANC) system, which will provide near real-time processing of drone airspace authorization requests nationwide. “With LAANC, the idea is to go from a 90-day approval process to one that takes just 90 seconds,” says Frank Matus, director of strategy and business development for aviation systems maker Thales Air Traffic Management U.S. The company is working with NASA, Syracuse University and others at the FAA’s New York State Griffiss International Airport test site to help develop the UTM, including LAANC. The FAA began testing a prototype LAANC system last November and plans to roll it out to nearly 300 air traffic facilities covering approximately 500 airports this year.

Sky’s the Limit

NASA’s UTM is a welcome development for Amazon CEO Jeff Bezos, who has hyped his company’s drone delivery service since December 2013 when he introduced the idea on 60 Minutes. Unwilling to wait for U.S. regulators to catch up with the company’s plans, Amazon in December 2016 made its first successful octocopter—propelled by eight small rotors—delivery of an Amazon Fire TV streaming media player (weighing about 280 grams) and a bag of popcorn to a home in Cambridge, England. Shipping businesses UPS and DHL are testing similar services.

Other companies are rivaling Uber’s flying taxi ambitions. In March Kitty Hawk—a start-up run by Google co-founder and Alphabet CEO Larry Page—introduced its Cora autonomous air taxi, which the company plans to launch first in New Zealand, although no official time line has been released. On January 31 Airbus’s Vahana single-person electric vertical takeoff and landing aircraft flew about five meters above a test site in Pendleton, Ore. The 5.8-meter long, 726-kilogram Vahana’s maiden voyage lasted less than a minute. About a week later Chinese drone maker Ehang released footage of its Ehang 184 quadcopter passenger drone making test flights. The company claims to have conducted more than 1,000 passenger test flights as high as 300 meters, as far as 15 kilometers, as fast as 130 kph and, at times, carrying more than 230 kilograms. The government of Dubai’s Road and Transport Authority in September held the country’s first public air taxi flight using an 18-rotor electrically powered prototype vehicle from German drone maker Volocopter.

Despite these successes, Embry–Riddle’s Ison is skeptical. “The idea of people owning or riding in passenger drones is like the dream of everyone having flying cars,” he says. “It’s cool that they have these [prototype passenger drones] they’re playing around with, but from a practical standpoint there are still a lot of areas that have to be addressed,” he says. Even delivery drones will have to wait until enough safety measures are in place. “I still think that what Amazon is envisioning is 10 years away if they’re lucky—meaning they don’t hit an airplane or kill someone.”

Even NASA’s Kopardekar acknowledges the prospect of drone deliveries in the next few years is up in the air. “Speculation is not a very good sport,” he says. “[Some people] are saying there’s a possibility to have point-to-point air taxis in certain locations, although not everywhere. Is that enough for economics? That’s a question that the individual companies will have to address, [although] there seems to be a market. Whether it’s possible technologically, that’s what we’re researching.”