Amazon Tests Delivery Drones At Secret Canada Site After US Frustration
An anonymous reader sends this excerpt from The Guardian:
Amazon is testing its drone delivery service at a secret site in Canada, following repeated warnings by the e-commerce giant that it would go outside the U.S. to bypass what it sees as the U.S. federal government's lethargic approach to the new technology. The largest internet retailer in the world is keeping the location of its new test site closely guarded. What can be revealed is that the company's formidable team of roboticists, software engineers, aeronautics experts and pioneers in remote sensing – including a former NASA astronaut and the designer of the wingtip of the Boeing 787 – are now operating in British Columbia. The end goal is to utilize what Amazon sees as a slice of virgin airspace – above 200ft, where most buildings end, and below 500ft, where general aviation begins. Into that aerial slice the company plans to pour highly autonomous drones of less than 55lbs, flying through corridors 10 miles or longer at 50mph and carrying payloads of up to 5lbs that account for 86% of all the company's packages.
Canada!
Amazon's getting smart with wheeled vehicles delivering to the big cities where flight is impossible, and flying devices to deliver to spread out communities like farm land...
The main problem (well, perhaps not the MAIN problem) I see is that no-one signed up to have drone flights right over their houses. You can buy and plan for where airports are going to be, but the "drone corridors" will just appear overhead one day. Drone sounds are (I think) especially obnoxious buzzing...
It'll be interesting to see if communities try to ban this.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
canada health care system is better for this as one some gets hurt they are covered and you are not facing bills racking up as the courts fight over who will pay for it.
There's only one way to punish Amazon for taking this activity outside of the US. We must find a way, since they have a business presence in the US, to add a larger regulatory and tax burden onto them until they submit, and return this activity, which we won't let them do anyway, to US soil. At which point of course we will not reduce that new tax or regulatory burden, but that'll show 'em anyway.
Way to go, Executive Branch.
Don't disappoint your bird dog. Go to the range.
The end goal is to utilize what Amazon sees as a slice of virgin airspace – above 200ft, where most buildings end, and below 500ft, where general aviation begins.
Too bad they are going to spend a significant time outside of that airspace during takeoff and landing. That airspace is also occupied by buildings higher than about ten stories.
It would be interesting to know the mean time between failure of the drones they intend to use.
https://xkcd.com/1243/
Given the rogue unregulated nature of this airspace they want to exploit, I will claim the airspace over my property as my own and setup a few, ummm, butterfly nets...
I live in Vancouver. The general attitude towards drones from both the public and the government is fearful and negative. The only reason they're getting away with this at all is probably because they're testing out in the woods somewhere.
As soon as drones are seen flying anywhere near populated areas, this whole thing will get shut down, fast.
Is this "slice of virgin airspace" within shotgun reach?
Try it! Library of Babel
.
Also, I saw no mention by Amazon about their concern for the safety of the people who will be living under the threat of the Amazon drones falling out of the air. Maybe that is why Amazon is moving so quickly, their concern for safety is not as the level it should be.
I'm having a hard time figuring out why I want hundreds of 50lb drones flying over my head just so Bezos can make another quick billion. There's no rushing when it comes to doing things the right way. I can't think of a right way to do it when one of these things fails and falls onto a freeway through someone's windshield, or lands on my neighbor's 2 year old playing outside in his sandbox. Maybe they've found corridors that don't cross over houses or populated areas, but I'm thinking that's unlikely. I'm all ears really, but how? Other than the fact that it's new and whizbang cool to have drones in the skies, there's going to be failures, and the fail cases for a car engine are far less disastrous than the fail cases for a drone over a certain size. Who knows, it may turn out that decreasing the amount of vehicular traffic actually decreases the number of traffic fatalities enough that drone failures are more pragmatic. But at least I'm able to actively minimize my time I spend exposing myself to the threat posed in traffic. With drones we're left up to whatever corridor path Amazon chooses. Happen to live underneath one of these? Tough luck.
Blackjack and hookers!
I am Slashdot. Are you Slashdot as well?
The article didn't distinguish much between testing stage regulations and production stage regulations. Did Amazon have problems finding a testing area? Why not just use the big backyard of some executive or some other private property to test?
But I imagine that regulations will be sticky in Canada also if and when they want to go to production deliveries. I've seen no evidence Canada has friendlier skies for real deliveries.
I suspect it's merely a PR stunt to embarrass the USA into creating friendlier sky laws.
Table-ized A.I.
If one falls and hits me or comes close, I'm keeping it and taking it home. Otherwise don't fly your drones over my head.
If the drone confines its flight path to mostly over the road systems it will make a lot less noise than a passing car.
I thought about that too, but the problem is road nose is well contained to buildings on the side of the street, while drone noise is elevated and thus can reach out a lot more.
Perhaps drone noise at 200+ feet would not be as bad as I'm thinking of, but it seems like these would be pretty large drones at 55lbs, thus quite a bit noisier than many of the drones we are used to hearing.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
On rainy and windy days, over farmland, during the holidays and in cities.
A little autonomous drones will be badly affected by wind and weather. Sure on sunny calm days they will work fine. Some stormy night and you want that tin of cavier delivered, and the drone will be wet and blown off course. What happens when they are struck by lightning?
Over farmland, the airspace starts at 0ft. Ag Aircraft are allowed to stay 500ft away from buildings and people, meaning over a random corn field, the aircraft can fly until their wheels touch. Bumping into a 50lb drone at 100kts will leave a dent in the aircraft, and probably destroy your package.
The world is mostly got paths clear around most houses, except during the holidays. A string of christmas lights may not be detectable by a drone, so what should it do when encountering it.
The cities are very dynamic. Delivery trucks, cats, dogs all change the terrain around your house. what happens when a delivery drone is being chased and caught by a curious cat. How about a large bird attacking it. Say the UPS driver just delivered a pile of packages to the front porch, and now there is no place for the Amazon drone to put something.
Then what about security? Certainly all a nefarious person would need to do drive around following drones, and collecting packages before the residents collect them.
How about people who live within 5 miles of an airport. Drones are not allowed there. No deliveries for you.
Yea, lots of work, for little payback.
... somewhere in vancouver BC, this much is confirmed
All I can say is: Skeet Shooting for Prizes!!!
Ok, I can understand Amazon doing their testing elsewhere, but I'm not quite sure what the resultant consequence to the US is.
When someone says, "Any fool can see
I wonder if the Drone will need Emissions controls.. then again even batteries outgas.. and what if they should collide with the Drones of another company.. who handles the Insurance claims?
Doc Brown and his Flying DeLorean never had it this good
The Stork101 models will Carpet Bomb customers.. just like Pidgeons.. Amazon is working out the logistics of tracking cell phone ground targets
I don't get this. What happens when the drone arrives at my address? Does it ring the doorbell and wait? Does it go round the back and leave it on the back porch if I'm out? (I'm in a low-crime neighbourhood where this is possible.) Or will it leave it with my neighbour, as instructed?
The airspace used by general aviation starts at the ground—not 500 feet. Crop dusters, helicopters landing and taking off at hospitals and private sites, as well as planes landing at airports all use the airspace all the way to the ground. In addition, it’s perfectly legal to fly below 500 feet as long as there are no people or buildings around.
I don't want fucking drones flying around my neighborhood making a bunch of noise and ruining my clear views of the trees and sky. Damn it!
That gives them a 200' altitude window (flight envelope) to play with. Let's take a closer look at this.
First, that altitude is AGL (Above Ground Level), so the flight envelope varies with the terrain. In hilly terrain this can quickly get complex, especially when what the FAA calls "obstructions" are taken into account. Obstructions include things like radio towers and tall office buildings. But the FAA is mostly concerned with obstructions that penetrate "controlled airspace" (above 400'), though their airchart database includes lower obstacles close to airfields.
There are areas where the 200' flight envelope can completely disappear, where no drone will be able to safely navigate. These will be "undeliverable" locations for Amazon. How will they determine where they are? 3-D knowledge of the low-level environment will be crucial, and the 3-D features of current satellite-based map services should be able to help.
What about other drones using the same airspace? In controlled airspace, the FAA requires all pilots to file a flight plan for non-trivial flights outside of "local" airspace. For long-distance flights, the FAA uses this information to assign planes to altitudes and routes. Will a similar centralized service be created to manage the narrow 200' drone window?
One more critical item of information is how well the drone will be able to keep to its assigned flight path. Wind gusts and thermals can cause sudden changes, so each flight plan will need to be separated from all others by enough distance to allow for course deviations.
Even with a central clearinghouse and coordinating center for commercial drones, the drones will have to cope with everything else sharing that airspace, including hobbyists flying RC aircraft and personal drones, not to forget kites, balloons and birds.
This means the drone must know:
1. Where "ground level" is.
2. Where the obstructions are.
3. Where "known" drones are.
4. What "other" things are in the immediate vicinity.
Even with all this knowledge, it's not enough: It fails to account for any failures or errors in the drone itself. Much more is needed.
First, the drone design must be very robust.
1. All flight-critical systems should have backups, including the following:
- IMU
- GPS
- Flight controller
- Engine controller
- Electronics power
2. The drone should be able to safely land with any of the following failures:
- A dead engine or lost propeller.
- A dead primary battery.
3. The drone should have the following safety systems in addition to the above backups and reserve capabilities:
- Obstacle sensors (radar, IR, ultrasound, video) to ensure the immediate airspace is clear.
- Parachute and/or airbags, automatically deployed, to minimize the risk of impacts.
- Manual flight remote control link (active in standby at all times).
- Active telemetry link
- Active video monitoring link (for use with above manual flight control and as part of obstacle avoidance system), preferably stereo and covering all directions (a spherical view).
I've assumed the drone will be electrical and operate from batteries: This may not be a commercially viable option (payload, time aloft, service time between flights), so additional safety concerns exist if liquid fuel internal combustion engines are used instead.
The real test case is for the drone to operate in a safe manner, or fail safely, even when hunters mistake it for a bird and try to perforate it.
Where are they planning to deliver to exactly?
They could be delivering to a closer distribution center or to the end customer?
If the delivery is to a distribution center then it is very inefficient given the small payload, if the delivery is tot he end customer then where would you leave the package? Drop it on the driveway where it would get run over by the owner? What about trees, power cables and other obstacles? Leave the package out in the rain? Would it land an leave the package on the ground or drop it an d potentially break the package.
Sounds like a good idea but I struggle to see how this would work
eh? Yeah ... what day is it (tomorrow) ? Humm.
It is standard practice for machinery safety, where live is at risk, for the system to be designed to both detect, and continue on to a safe stop, when a fault occurs. It's all based on probability of failure, frequency of exposure, consequence of failure. Also people are jumping to a lot of conclusions. I work in postal industry and we are waiting for self driving cars that launch drones to do the 'final metres' to the front door. That appears to be the ultimate minimum price (road cheaper than air and drone cheaper than robot).The idea of drones flying long distance doesn't seem to make any economic sense - they can only take one light parcel with them. A system with auto car and drone could get delivery of parcels cheaper than cost of current stamp for letters, easily.
Yes, the R&D and testing have to happen before this will be economical and safe. Traffic management and liabilities also need to be worked out. So let that happen and stop the "what if's" and "it isn't safe yet" and "it is impractical" ect. The FAA needs to be reasonable and let this develop in a safe manor, not just lock it down so it never gets anywhere.
Also remember that the FAA already allows up to 55lbs to fly with very little restrictions if it is NON-COMERCIAL AKA unlicensed pilots, and little more than civil accountability for what happens if there is a crash.
Why not just use the big backyard of some executive or some other private property to test?
Because, as the article stated, the FAA is dragging its feet on granting them a license to do that.
I see RC planes from time to in fields and parks. Are drones considered different? What's the Fed weight limit for an RC plane?
Table-ized A.I.
Sounds like they're just as dishonest as Iran - secret buried bunkers. It looks like it's time for Canada to get some sanctions!
So I doubt I will be getting anything delivered by drone.
I am very small, utmostly microscopic.
I honestly never thought this idea would... Take off!
The FAA has been suing anybody commercially flying drones anywhere, with very broad definitions of commercial. If you take a video of your house from a drone hovering 3 feet off the ground and then put it on youtube with ad revenue enabled, the FAA will go after you if they notice it.
The FAA did agree to let Amazon test drones in the US, but with a number of restrictions including having a licensed pilot, keeping it in line of sight, only using pre-approved models, etc.
(that's Area 51 in American)
"Win treats sysadmins better than users. Mac treats users better than sysadmins. Linux treats everyone like sysadmins."
The end goal is to utilize what Amazon sees as a slice of virgin airspace â" above 200ft, where most buildings end, and below 500ft, where general aviation begins.
Yea, I hate to break it to Amazon, but that airspace isn't "virgin", it is currently the domain of helicopters.
Normally, helicopters will fly 500ft above the ground, give or take a bit, but they do not have a lower limit.
FAR 91.119 says "Except when necessary for takeoff or landing, no person may operate an aircraft below the following altitudes..." and it turns out to be illegal to fly an airplane less than 1000' above the rooftops of a city (i.e., about 1200' above the ground) or 500' from any person, vessel, vehicle, or structure in the countryside (i.e., at least 500' above the ground). This is a much closer look than you would get in a commercial airliner, but it isn't all that close. FAR 91.119(d) says "Helicopters may be operated at less than the minimums prescribed in paragraph (b) or (c) of this section if the operation is conducted without hazard to persons or property on the surface."
As long as the helicopter could autorotate to a tennis court, road, or field in the event of an engine failure, the pilot can fly much lower than in an airplane.
Source - Me... I'm a Certified Flight Instructor in both helicopters and airplanes, with thousands of hours of dual given, 2 years spent as a Chief Flight Instructor in a FAA Part 141 school, as well as nearly 10 years of commercial flight experience in tours, offshore, and EMS.
Amazon is out of their mind if they think this is ever going to work. If nothing else, the police and EMS helicopters are not going to get out of their way and they have more of a need to be there.
Looks like April 1st
These are not the drones you are looking for...
"
You forgot about the ease of delivering that 5 pound block of C4 plus detonator to pretty much anybody that ordered it. It isn't even "just" the Amazon drones. Anybody can capture an Amazon drone (or build their own copy and paint it accordingly) and use it to make a "special delivery" to, well, pretty much anyone. "Special Delivery, Mr. President! It's those "books" you ordered from Amazon!"
You can pack a whole lot of evil into 2 kg of C4 (or whatever the latest/greatest compact explosive is) plus detonator. You can saturate any reasonable defensive system by having 100+ drones attempt a delivery at the same time. You can carpet bomb crowded marketplaces - the drone itself will conveniently supply the shrapnel, or you can fly the drones under cars or into glass-front buildings before detonating. And best of all, you can do it in complete anonymity and safety! The drones will be impossible to track back to a point of origin, flying literally under the radar and in numbers too great to track anyway. You can rent a barn or warehouse, ship in as many amazon-a-likes as you can, load them with Sarin, with Anthrax, with weaponized Ebola or with powdered radioactive waste, or -- what the heck -- with all of these at once, to saturate and overwhelm even emergency response systems with multiple distinct threat vectors, and after launching them with a program that directs them to converge on a given target from all directions after initially moving on "delivery" trajectories to a spread of locations, "
None of your comments apply to any regulations postulated by the FAA. You are confusing dangerous possibilities from people with criminal intent with beneficial uses by responsible parties. It is not in Amazon's interest to accrue liabilities from drone crashes. They will be sued for their back teeth and they know it. On the other hand, criminals will not follow FAA regulations whatever they are.
It is perfectly reasonable to allow Amazon the time and space to work out any glitches in a safe region of the US.
It's also perfectly reasonable to assert "That is never, ever, going to be legal, so why bother to permit then to develop something we aren't going to allow." Admitting that this isn't quite what has been said, it is probably at the heart of the denial.
As you suggest, there will probably be little we can do about preventing people from doing this with drones already and/or in the near future -- the barn door already done been opened and the horses are long gone -- but at least we can arrange a world where ANY drone, autonomous or not, is illegal and hence a candidate for defensive measures.
It is, sadly, yet another case of the tragedy of the commons. Drones could be everything from fun to enormously useful, but once they are in "the commons" and available to everyone, it only takes a few butt-holes to make them a liability that exceeds any possible benefit. At the same time, it is virtually impossible to put techno-genies back into the bottle. I can hardly wait for the day somebody figures out how to make a compact 20 kt+ nuclear device out of commonly available materials and using nothing but garage-shop tools. There have been a very few SF stories written about such things, and none of them have happy endings.
Maybe this is the answer to the "Where are they?" paradox. Any time a civilization rises above a certain technological threshold, it becomes possible for anybody to destroy it, and there is always somebody that does. Maybe it becomes possible to build a fusion device that doesn't require a fission trigger, for example, or maybe it becomes possible to bio-engineer a doomsday virus with a "perfect" epidemiological profile and nearly 100% mortality in your basement.
Fully anonymous drone-delivered murder and mayhem may not qualify as doomsday, but it will certainly create some serious challenges to the concept of personal freedom vs public safety and the abuse of the commons.
rgb
Even when the experts all agree, they may well be mistaken. --- Bertrand Russell.
They can be made safe without government regulation. They can carry transponders. They can carry radar reflectors and show up on airport radars. Over time I'm sure regular flyways will be established, the same as footpaths and roadways were in the past. These can be mapped to 3D GPS. All of this is current technology.
The trick is NOT killing every possible new tech because of some imagined danger, but ironing out the bugs. That is exactly what Amazon wants to do.
By the the standards you claimed, we would not have air travel because of the possibility of 9-1-1.
So here I am, building my death drone fleet, thinking "Gee, I'll be sure to leave in the transponder and radar reflectors. After all, even though there are no government regulations requiring them, I want my victims to see them coming."
As for 911, a) the economic impact of air travel and freight is difficult to equate to the net productivity of a small-payload drone fleet at any level; b) the "solution" to the problem turned out to be literally locking the pilots into the front of the plane so passengers would have to cut through a solid bulkhead in order to reach the cockpit, which would at the very least require time and make the enterprise not worth the risk. There might be more measures we don't know about as well. This is possible because airplanes are an enormous capital investment in addition to being a core feature of our economy in ever so many ways. Adding security features to an airplane doesn't double its cost or halve its productivity.
With a drone, that is not the case. Security measures added to a drone do significantly increase its cost and lower its productivity (which is probably marginal anyway, given its small payload). And there is no way that I can imagine to add security features that somebody can't just turn off or remove or trivially suborn without massive government intervention.
It's sort of like the various assault rifles being sold that can have a single part or two removed or replaced and they are magically transformed from being "semi" to "full" automatic. Marvelously effective at preventing honest citizens from owning fully automatic assault rifles, not so good at preventing full automatic assault weapons from showing up all over the place in the hands of the less honest or fringe militias.
But hey, now the fringe militias will be able to add another weapon to their repertoire! Full automatic rifles, vans full of homemade TNT and an explosives-laden drone fleet. I can't wait!
And yes, we'll get this in spite of the government trying to slow it down. I'm betting we'll first see it in drones built and delivered by personality disordered borderline high school students who by some means manage to lay their hands on real explosives. That sounds like so much fun I'm tempted myself...
rgb
Even when the experts all agree, they may well be mistaken. --- Bertrand Russell.
Experienced modeler or FAA pilot without modeling experience, which do you want flying them?
I've been flying radio control model airplanes for 40+ years and helis for 25 and quadcopters (fly the same as helis) for less. I have seen many a certified FAA pilot bring his shiny new toy RC airplane to the field and crash it, all the time claiming that he knows what he is doing.
Well he doesn't. When the controls reverse because he is flying towards himself (left is right and right is left) he crashes. He doesn't know about batteries, radios, engines, model airworthiness or a host of other things.
Don't take this as bragging. It is just fact. I can and do fly RC airplanes an entire flight of continuous rolls, with either left handed or right handed transmitters (mode 1 or 2 actually). I actually write code for my quadcopter flight control board, including stabilization and GPS guidance. If something goes wrong, my reflexes know to do the right thing. You think an average FAA pilot knows anything about any of this?
Yet he can get a license to fly a quadcopter commercially and I can't. You only need to know 1% (my opinion, not researched) of an FAA license to handle a quadcopter legally, yet you need to know 100% of RC quadcopter modelling to fly it safely.
WHICH DO YOU WANT FLYING OVER YOUR HOUSE?
FAA is an old man government dynasty that is trying to protect itself. By it's own formulas (and this is good math), it is probability of failure times expense of failure that makes guidelines for reliability. By that guideline, a 750,000 lb aircraft, traveling at 500mph needs to be 1,000,000 times more safe (ratio of weight times squared ratio of speed) than a 7.5 lb multicopter flying at 50mph. Another way to look at it is that it takes 1,000,000 such multicopter crashes to equal 1 FAA airplane crash. The number is even higher than that because every destructive FAA crash kills people, where very few destructive multicopter crashes kill people.