Sunday, August 23, 2020

To Build a Plane That Can Fly Itself, Start With a Pilot in the Cockpit: Startup Xwing envisions testing automated controls on cargo planes

The Wall Street Journal
By Andy Pasztor
August 20th, 2020 12:28 pm ET

A Northern California startup has devised a novel strategy to create planes that can fly themselves, by first testing a new generation of automated flight controls during conventional cargo flights flown by pilots.

Starting with videos of a demonstration flight slated for release Thursday, San Francisco-based Xwing Inc. aims to jump-start such technology by showing its reliability on current turboprops already flying established routes in everyday conditions. Relying on a handful of specially outfitted Cessna Grand Caravans—workhorses of short-haul, single-pilot cargo operations nationwide—the company seeks to gather data about navigation, automated emergency maneuvers and autonomous landing features ultimately intended for commercial use without anyone in the cockpit.

Proponents say U.S. authorities could authorize selected flights without pilots over stretches of water or uninhabited regions as soon as the first half of 2022, potentially years before the Federal Aviation Administration certifies all-new, comparably sized autonomous designs being developed by competitors. In addition to the person flying the plane, pilots on the ground are slated to monitor trips and communicate with air-traffic controllers, at least initially. The single-engine planes carry some 4,000 pounds of cargo, and Xwing anticipates cruising at altitudes under 10,000 feet and at 230 mph—well below most airline operations.

“This is the fastest way to get to commercial uncrewed flight,” said Marc Piette, the company’s founder and chief executive. By tapping into existing markets, he said, “it’s our way to basically get the regulators on board” with the concept. Xwing’s technology, which also can be adapted for use on other conventional aircraft, is part of a National Aeronautics and Space Administration initiative aiming to introduce self-flying vehicles into U.S. airspace.

If demonstration flights roll out without major hitches and validate the company’s approach, even limited regulatory approval is still years away. Steps to phase pilotless aircraft into U.S. skies and traffic-control procedures are bound to take longer, as startups and major aerospace suppliers flock to invest in proposed urban flying taxis and an array of drones.

Without commenting on specifics, a Federal Aviation Administration spokeswoman said “the time it takes to certify an aircraft depends on many factors, including the magnitude and complexity of the proposed product, the applicant’s maturity and the reliability of its systems.”

Established companies with much deeper pockets, including Boeing Company are targeting some of the same customers with new electric-powered prototype aircraft. Google, Amazon Inc and United Parcel Service Inc are intent on winning regulatory approval for much smaller drones, intended to eventually deliver packages weighing several pounds directly to consumer doorsteps.

Simultaneously, General Atomics and other companies are looking to convert high-altitude military surveillance drones to law enforcement and other civilian applications. Venture funds and other investors are supporting a throng of startups around the globe developing autonomous flying vehicles capable of carrying a handful of passengers looking for alternatives to traffic-choked streets and highways.

Several private turboprop and personal-jet models already have automated systems able to navigate and land planes, without any human input, in the event of a pilot suddenly experiencing a medical emergency.

Some Wall Street analysts and industry experts, though, are convinced that unglamorous cargo operations similar to those Xwing envisions—particularly focused on what are called feeder routes that stretch a few hundred miles—are likely to become pioneers of widespread, fully autonomous flights.

Regulatory hurdles and public concerns, not technological limitations, are the primary reasons self-flying cargo planes aren’t already in use, according to Kevin Hiatt, a former high-ranking airline and industry safety official. “Technology has advanced fast enough and far enough already,” Mr. Hiatt said.

Jia Xu, a senior strategy official at Honeywell International Inc told an aviation panel last month: “We have a vision that the future of air cargo will be completely autonomous.” In recent months, Honeywell has established an unmanned aircraft unit partly focusing on cargo operations and created a simulation lab to examine human-machine integration.

Rather than designing brand-new flying machines, cramming them with automated systems and seeking regulatory approval for the entire package, Xwing’s approach is incremental. The Caravan’s structural safety, engine performance and overall reliability already are established. So are many of the routes and smaller airports served by the planes. Fueling and maintenance infrastructure likewise is assured. Xwing, which is partly funded by France’s Thales SA hopes to build on that pedigree to reduce the time and complexity of regulatory vetting.

The Federal Aviation Administration has issued experimental licenses for Xwing, while approving installation of various aircraft-detection and avoidance sensors and related software.

Jay Merkle, head of the Federal Aviation Administration drone-integration office, has stressed that autonomy should be primarily a tool to make flying safer and open up new commercial applications. “It’s about finding those places where human error is making the safety system very fragile,” he told the same industry panel in July.

But once drones and larger autonomous aircraft proliferate, Mr. Merkle added, the Federal Aviation Administration and foreign regulators must ensure automated air-traffic control networks keep pace. There is “absolutely no way,” he said, “we could hire enough controllers or field enough equipment to manage” the extra traffic using current procedures.

Original article ➤

1 comment:

  1. Another "the future of aviation" promo video, but somehow nobody involved in the project recognizes the error of showing a bank angle reaching almost 90° during that turn at the end.

    Impress pilots and regulators by disregarding normal category certification for angle of bank not exceeding 60°. Did the breakthrough controls engineer really want to convey that the automation would routinely make a turn like that?