Amazon Prime Air drones will deliver packages weighing up to 5-lbs using robotic drones like this one. (Click Image To Enlarge)
On a 60 Minutes segment last night, Amazon CEO Jeff Bezos revealed a plan that would put all other rapid shipping options to shame: a fleet of little drones that could carry items from a warehouse conveyor belt to a buyer’s doorstep, all within half an hour.
Hardware limits, safety questions, and FAA rules are all keeping Jeff Bezos' dream from becoming a reality
Like Google’s driverless cars or, on a grander scale, Elon Musk’s Hyperloop, the project — dubbed Amazon Prime Air — seeks to revolutionize how things and people move, offering speed, convenience, and the futuristic experience of a flying octocopter postman. Suggesting it could be possible in four to five years, Bezos said.
"I know this looks like science fiction. It’s not."
But in order to turn its ambitious concept into a reality, Amazon has a lot of work to do.
Amazon has said that its unmanned aerial vehicles [UAVs] should be able to carry a 5-pound package up to 10 miles within 30 minutes. To do that, it’s going to need a lot of battery power. Octocopters on the market today are often built to carry cameras heavier than 5 pounds, but they fly only a fraction of the distance; the SteadiDrone EIHG8, for example, is meant to carry between 2 and 13 pounds, but only for a maximum of 15 minutes and less than 1 mile — the heavier the package, the worse the performance. Aerial robotics company Skycatch, which has been working on a similar drone delivery project, sees the solution as swapping batteries automatically throughout the flight while traveling in an airspace corridor dedicated to unmanned vehicles.
Missy Cummings of MIT’s Aeronautics and Astronautics program thinks an Amazon drone would require a big leap in technology. She says.
"Unless Jeff [Bezos] has made some kind of amazing breakthrough in battery life, which I kind of doubt, the flight time is going to continue to be a problem."
Some have speculated that Amazon is building off hardware and software from Chris Anderson’s UAV company 3D Robotics, pointing to a GPS and compass module in the concept video; 3D Robotics declined to comment on this possibility. If Amazon is able to build a strong enough battery and account for variables like wind, the octocopters could fly semi-autonomously with a human supervisor, who Cummings suggests could manage up to 20 or 30 of them at a time, though the FCC’s rules are still largely unknown.
Once you build an octocopter that can make a 10-mile trip in half an hour and create a system to manage it, you start getting to the really interesting questions. How, for example, would you stop people from taking down Prime aircraft and stealing their cargo? Ryan Calo of the University of Washington’s tech policy lab. There’s also a more low-tech option: shooting the drones. Cummings says.
"Instead of shoplifting, we may begin to see delivery hacking. If you saw a drone flying fairly close overhead, maybe you could whip out a gun. Maybe you could whip out a slingshot. They would have to stay high long enough to basically either be out of sight or not in range of a weapon."
While protecting its drones from hijacking, Amazon has some major safety concerns to work out. Any plane can crash, and the octocopters wouldn’t have the same possibility of pilot error, but they'd also do something risky and virtually unprecedented: drop from a potential height of a few hundred feet, possibly with almost no human direction, right onto your driveway. Or your herb garden. Or, if you’re particularly unlucky, your dog. The sensors that keep planes from hitting each other in midair aren’t built to avoid tiny obstacles on the ground, though Amazon could use semi-experimental lidar systems like the ones on Google’s driverless cars to avoid obstacles. An Australian company could provide a test case — called Zookal, it plans to let urban buyers order textbooks and then wait outside for a drone to drop its package while hovering above the ground. Since its pilot program won’t start until next year, though, it’s hard to say how well this plan will actually work.
One potential answer is to combine this program with something like the Amazon Lockers deployed at 7-Elevens and drugstores in a few test cities. In a blog post, Skycatch describes what Amazon needs to make Prime Air a reality, including safety and anti-theft measures. After making most of its journey high in the air, it says, an ideal UAV would drop inside "authorized pick-up and drop-off stations" that would prevent theft and reduce the chances of a bad landing. It’s not as charmingly futuristic as a machine delivering something right to your door, though it’s still a lot more convenient than traditional shipping, and it’s probably what we’ll get in the foreseeable future. Skycatch’s Christian Sanz says.
"No one is going to fly a drone like the [Amazon] video suggests; too many things can go wrong. Props can take someone's eye out, batteries don't last long so you need an automated way to swap them, power issues are the main cause of UAVs falling off the skies, so you need a way to manage that. If the UAV falls from the skies you need a way to soften the landing."
All of this, of course, is making the vital assumption that the FAA will allow the program in the first place. Amazon has been vague about when exactly we’ll see Amazon Prime Air deployed; an FAQ says that it "will be ready" when the FAA’s deadline for commercial UAV guidelines comes up in 2015. But although the FAA is making slow progress towards figuring out how to certify UAVs and their operators, it hasn’t even completed a basic, early step: officially approving a set of six flight ranges to test its rules. From there, it’s going to slowly ramp up a pilot program, and by 2015, it’s not required to do much more than tell companies how to responsibly fly a commercial UAV and help them start testing. Amazon might well be ready at that point, but "ready" won’t mean one-click drone shopping.
Amazon is also taking aim at one of the most sensitive regulatory issues: letting drones fly above our cities and suburbs. Despite all those great cityscapes that filmmakers capture with UAVs, even hobbyists aren’t supposed to fly in populated areas; if you want justification, look no further than the quadrotor that dropped a few dozen stories and barely missed hitting a Manhattan businessman this fall. While this raises problems for Amazon, though, having a huge company take on the issue is probably good news for other UAV proponents. The company is a lobbying heavyweight in the tech sector beloved by American politicians, and it’s all but certain to push the FAA for faster rulemaking and more permissive policies, as long as it can demonstrate that looser rules won’t pose a significant safety problem. Whether this will actually speed the FAA’s decision is a matter of debate.
For the rest of the commercial drone world, Amazon Prime Air is a huge publicity coup. At best, small UAVs have been seen as potentially helpful novelties in the past; at worst, they’re associated with ubiquitous surveillance or the much larger military craft responsible for bloody airstrikes in places like Yemen and Pakistan. Now, though, Amazon is promoting its octocopters as a convenient and futuristic shipping option for its massive user base. For anybody with serious privacy or safety concerns, meanwhile, the plan is an incentive to speed up work on meaningful regulations. Senator Ed Markey (D-MA), for instance, has used the Amazon announcement to promote his UAV privacy bill. He said in a statement,
"Before our skies teem with commercial drones, clear rules must be set that protect the privacy and safety of the public."
If anyone’s drones are in a position to teem, Amazon’s are. In 2012, the company said it shipped 15.6 million items on its busiest holiday shopping day; based on Bezos’ claim that 86 percent of packages are under 5 pounds, that could add up to a peak of 13.5 million eligible shipments per day, a number that will likely only grow over the years. Amazon’s FAQ says.
"One day, Prime Air vehicles will be as normal as seeing mail trucks on the road today."
But a sky swarming with yellow octocopters is still far, far away, especially in the US. Cummings says.
"There will never be a time in our foreseeable lives where we look up in the sky and see a bird-trail of UAVs flying back and forth. I suspect that within the next 10 years, you will occasionally see one going by."
Within five years, she thinks the FAA might let Amazon send delivery vehicles between approved fulfillment centers; for anything like Amazon’s door-to-door concept, more like 10 to 15 if it happens at all.
That doesn’t mean, however, that we might not be seeing Prime Air drones in the sky sooner — it just won’t be over American soil. Besides Australia’s Zookal, Chinese company SF Express is reportedly testing a delivery UAV, albeit only in remote areas. Cummings says.
"Amazon is not stupid. They’ll go to other countries."
Bezos, meanwhile, has made clear that his company is in this project for the long haul. He said in 60 Minutes.
"I don’t want anybody to think this is just around the corner. This is years of additional work from this point."
But it will almost certainly take more than his optimistic estimation of a half-decade before the drones arrive at our doorsteps.
COMMENTARY: I remember what people were saying when Jeff Bezos announced that Amazon would be producing the Kindle, it's highly successful book eReader. Then he announced that Amazon would be producing its own tablet, the Kindle Fire. Well, folks. Never underestimate Jeff Bezos. The guy is a genius when it comes to developing great new ideas and products, and I would take Amazon's Prime Air robotic drone delivery service very seriously. Obviously a lot of kinks still need to be worked out, and the FAA has to establish guidelines for commercial UAV's like Prime Air, so I don't see this happening for several years. Having said this, I can definitely see a company, if not Amazon, using commercial UAV's to make deliveries of goods or provide services.
Courtesy of an article dated December 2, 2013 appearing in The Verge