Why Amazon’s drone plan may not get off the ground

A test of an Amazon delivery drone.

A test of an Amazon delivery drone.

This weekend Amazon Chief Executive Officer Jeff Bezos told 60 Minutes that he wants small unmanned aerial vehicles — drones — to speed packages to online shoppers as early as 2017, cutting delivery times to as quick as 30 minutes.

It’s a bold, imaginative plan — one that could propel a host of technological and legal advancements.

It’s also really, really difficult to pull off. What follows are just four of the reasons Bezos’ Amazon delivery-drones might not get off the ground.

1)  Drone delivery flights are illegal, at present. Among other prohibitions, the Federal Aviation Administration bans drone flights over 400 feet altitude and near airports and populated areas. Bezos’ plan is for the robots to take off from fulfillment centers near big cities. They might be able to stay below 400 feet and avoid airports. But exactly how can a drone deliver a package to a populated area without flying over … a populated area?

More important, the FAA currently bans all commercial uses of drones. Simply stated, you’re not allowed to make money off them — which is exactly what Bezos aims to do.

The good news for Amazon is that Congress has required the FAA to loosen its drone restrictions. The agency has said it will roll out new rules in mid-2014. “That will really be first time anyone can fly an unmanned aerial system for hire,” says Ben Gielow, general counsel for the Association of Unmanned Vehicle Systems International, a drone trade group.

“But looser laws are only a first step,” Gielow adds. “The regulations that govern operator certification, airframer certification, how maintenance is done, how the communications links work and all that stuff all need to be worked out.”

2) Drones are expensive. Bezos showed 60 Minutes some prototype delivery drones. They’re “octocopters,” named for their eight helicopter-like rotors. Octocopters and their four-rotor cousins called “quadcopters” are among the most popular unmanned vehicles in use among universities, corporate laboratories and non-military government agencies.

But they’re not cheap. A high-performance drone — one capable of long-range flight at high speed while also carrying several pounds worth of packages — can set you back $50,000. Middling models are around $3,000. Budget drones such as the $300 AR Parrot are notoriously flimsy and unreliable — and lack the horsepower to haul more than a few ounces of payload.

Based on my rough calculations, a FedEx driver and his truck cost $40,000 a year to acquire and operate — including the truck’s purchase price spread over a decade of use. A typical driver can deliver 75 packages a day. It would take at least six drones costing $300,000 to do the same amount of work, assuming the robots always work perfectly. The drones would have to fly 12 hour-long, round-trip deliveries five days a week for eight years in order to be even a dollar cheaper than a human driver.

Eight years is a long time for a tiny, complex machine prone to crashes and malfunctions. The bottom line is that people are probably cheaper.

3) Drones are dumb. People can read maps, follow directions, navigate lawns and foot paths, step over shrubs, squeeze onto cluttered porches, read house numbers and ring doorbells to let customers know their packages have arrived.

By contrast, even the most sophisticated flying robots lack the ability to read numbers and dodge obstacles such as wires and birds. There’s no way they can ring your doorbell. The best today’s octocopters can manage is to follow GPS coordinates to the approximate location of your house, accurate to within a few yards. But the machine will have no way of knowing where on your house the front door is — to say nothing of finding the doorbell.

“In order to do things useful for people, the robot has to know its environment,” said Stefanie Tellex, a roboticist at the Massachusetts Institute of Technology. Tellex and other researchers are hard at work on smarter drones. By equipping the robots with laser sensors and large databases of known objects, they hope drones can quickly deduce the layouts of new buildings and landscapes and maybe pinpoint the door.

But the technology is not very far advanced — and might not even be ready in the five years before Bezos wants to launch drone delivery.

4) People distrust drones. Many of the world’s most sophisticated flying robots are in the military’s possession. They fly high over foreign battlefields, spying on suspected insurgents and terrorists and even attacking them with missiles and bombs. In the United States, the Department of Homeland Security is a major drone user.

The use of flying robots in warfare and surveillance has instilled a deep suspicion of drones in most people. This year the small town of Deer Trail, Colorado, even tried to make it legal for residents to shoot down any drone they saw overhead — a move that elicited a stern rebuke from the FAA.

But the good people of Deer Trail aren’t entirely off base. The National Security Agency is known to collect data on Americans’ Internet activity and phone conversations, either by cooperating with telecommunications companies or by secretly slipping spy software into private and corporate networks.

Could federal agencies resist the temptation to hijack Amazon’s fleet of drones as they zip back and forth over much of the American population? And do people want 30-minute delivery so badly that they’re willing to risk it?

David Axe is the national security editor at Medium.com. He has written for Danger Room, Wired and Popular Science. He most recent graphic novel is “Army of God: Joseph Kony’s War in Central Africa.”

_______________________________________________________

Copyright  © 2013 Capitol Hill Blue

Copyright  © 2013 Thomson Reuters  All Rights Reserved.

Enhanced by Zemanta

3 Responses to "Why Amazon’s drone plan may not get off the ground"

  1. Pondering_It_All  December 4, 2013 at 10:27 pm

    Finding the location to leave the package would very simple: Just require customers who want drone delivery to place a landing pad in some safe location in their yard. The drone could locate it by activating a little Bluetooth radio beacon that sits under the center of the target, when it gets within a few hundred yards. As for ringing the doorbell, that’s just dumb: The drone would be in contact with the delivery system computers at all times, so it could email or text you when your package is on the ground.

    The real problem would be safety: Like a toddler or a pet cat walking right into one of those spinning propellers and slicing off a body part!

    Considering the weight limitation of these drones, a better comparison would be a bike messenger or pizza delivery. A regular UPS truck makes no sense for 30 minute deliveries, since it holds an entire day’s worth of delivery items. Sending drivers out in little electric cars makes more sense than trying to send a drone for each item. Each driver would get his load at the distribution center and have his GPS programmed with the best route at the same time. The one thing Amazon absolutely excels at is algorithms like optimal packing, so each load would fit perfectly into the driver’s car! If this makes economic sense for pizza delivery, it has to also work for similar small items, with customers willing to pay for quick delivery.

  2. woody188  December 5, 2013 at 6:50 am

    The whole thing is a hoax/publicity stunt. A better design would be a small hydrogen dirigible with foam propellers as it would improve safety 10-fold, but helicopter style drones are all the rage.

  3. Joe  December 5, 2013 at 11:38 am

    Would the biggest and baddest delivery dirigible be named “The Nerf Zeppelin”?
    Will Bezos start a public information campaign to educate the public to sing, in four-part harmony, “Oh, the humanity!” whenever one goes up in a hydrogen-fueled fawhoomp?

    These are important questions that need answers!

Comments are closed.