Amazon Is Still Struggling To Make Drone Deliveries Work (theverge.com) 69
An anonymous reader quotes a report from The Verge: A report from Bloomberg details the obstacles hampering Amazon's efforts to get its delivery drone program off the ground, citing a high employee turnover rate and potential safety risks. According to Bloomberg, there were five crashes over the course of a four-month period at the company's testing site in Pendleton, Oregon. A crash in May took place after a drone lost its propeller, but Bloomberg says Amazon cleaned up the wreckage before the Federal Aviation Administration could investigate. The following month, a drone's motor shut off as it switched from an upward flight path to flying straight ahead. Two safety features -- one that's supposed to land the drone in this type of situation and another that stabilizes the drone -- both failed. As a result, the drone flipped upside down and made a fiery descent from 160 feet in the air, leading to a brush fire that stretched across 25 acres. It was later put out by the local fire department.
Last year, a Wired report revealed that Amazon's drone delivery operation is struggling just as much in the UK, despite making its first-ever drone delivery near Cambridge in 2016. Wired's report suggests that the UK outfit is marred by some of the same issues described by Bloomberg, including a high turnover rate and potential safety issues. At a UK-based facility for analyzing drone footage for people and animals, one worker reportedly drank beer on the job, while Wired said another held down the "approve" button on their computer regardless of whether there were hazards in the footage or not.
Former and current employees at Amazon also told Bloomberg that the company is prioritizing the rushed rollout of its drone program over safety. Cheddi Skeete, a former drone project manager at Amazon, said he was fired last month for speaking with his manager about his safety concerns. Skeete told Bloomberg that he was reluctant to continue testing a drone that had crashed five days previously but was told the team had inspected 180 engines on 30 different drones -- Skeete doubted this assertion, as checking the motors is a cumbersome process, Bloomberg reports. David Johnson, a former drone flight assistant for Amazon, told Bloomberg that Amazon would sometimes perform tests "without a full flight team" and with "inadequate equipment." Johnson also said the company often assigned multiple roles to one person, a claim Bloomberg says is corroborated by two other former Amazon employees. "They give people multiple things to do in a very narrow window of time to try to boost their numbers, and people cut corners," Johnson told Bloomberg. "They were more concerned about pumping flights out and didn't want to slow down."
Last year, a Wired report revealed that Amazon's drone delivery operation is struggling just as much in the UK, despite making its first-ever drone delivery near Cambridge in 2016. Wired's report suggests that the UK outfit is marred by some of the same issues described by Bloomberg, including a high turnover rate and potential safety issues. At a UK-based facility for analyzing drone footage for people and animals, one worker reportedly drank beer on the job, while Wired said another held down the "approve" button on their computer regardless of whether there were hazards in the footage or not.
Former and current employees at Amazon also told Bloomberg that the company is prioritizing the rushed rollout of its drone program over safety. Cheddi Skeete, a former drone project manager at Amazon, said he was fired last month for speaking with his manager about his safety concerns. Skeete told Bloomberg that he was reluctant to continue testing a drone that had crashed five days previously but was told the team had inspected 180 engines on 30 different drones -- Skeete doubted this assertion, as checking the motors is a cumbersome process, Bloomberg reports. David Johnson, a former drone flight assistant for Amazon, told Bloomberg that Amazon would sometimes perform tests "without a full flight team" and with "inadequate equipment." Johnson also said the company often assigned multiple roles to one person, a claim Bloomberg says is corroborated by two other former Amazon employees. "They give people multiple things to do in a very narrow window of time to try to boost their numbers, and people cut corners," Johnson told Bloomberg. "They were more concerned about pumping flights out and didn't want to slow down."
Re: (Score:1)
I’m happy you’re at least updating your poor taste autist bullshit.
Re: (Score:1)
"SJW" Slack jawed whites? WTF?
Re:Dumb Idea Is Dumb (Score:5, Interesting)
It's the kind of thing that could give Amazon a huge advantage for at least a few years, before rivals managed to develop their own technology. It could also really hurt Amazon if one of their rivals got there first, so they are obliged to work on it.
Sounds like they are being cheap though. The drone video review sounds like minimum wage contractors. You can't trust them to do anything, you have to verify every action was completed properly. For stuff like this you can't use them at all. Uber discovered the same thing with their autonomous vehicle safety drivers.
Re: (Score:1)
Because "drone deliveries" is a stupid and unworkable idea.
Clearly you don't remember using or weren't around for old versions of Windows and Mac OS. Those crashed all the time too.
Drone deliveries make a lot of sense for shipping things that aren't profitable enough to have a human deliver, like when your smoke alarms start beeping at 3AM and you really need a few 9 volt batteries.
Re: (Score:2)
Re: (Score:1)
It was just an off-the-cuff example. I'm sure some people do keep a stash of batteries squirreled away. I personally don't, because they're generally dead by the time I need to use them.
As for the noise, that depends on how quiet your neighborhood usually is. I suppose for people who live in neighborhoods without loud car exhausts, excessively subwoofer'ed car stereos, and police helicopters circling every night, drones dropping off Amazon packages might get annoying. Must be nice to live in that kind o
Re: (Score:2)
Don't smoke alarms warn you about the low battery a couple of months in advance?
And: Isn't is something that could wait until morning? Just switch the damn thing off for a couple of hours.
Re: (Score:2)
Re: (Score:3)
Why would you need to replace your batteries at 3 AM? Pull the battery out, go back to sleep, pick up batteries during the day when you're out.
You're right, that was a one off example and a poor one. But here is one. You live in the sticks in Montana where you have a two mile driveway. The dro
Re: (Score:2)
Still not a particularly useful scenario. Useful would be if a blizzard just deposited 10 feet of snow in Montana and you won't be able to dig out for weeks, and the drone can deliver something to your door.
Re: (Score:2)
Still not a particularly useful scenario. Useful would be if a blizzard just deposited 10 feet of snow in Montana and you won't be able to dig out for weeks, and the drone can deliver something to your door.
Like a handgun and two bullets?
Oh, wait: Everybody in Montana already has that.
Re: (Score:2)
Because "drone deliveries" is a stupid and unworkable idea. Just another example demonstrating that having a lot of money doesn't automatically make you smarter.
I completely agree. There is no point in this and no reason for the public to accept unnecessary risks associated with quad copters buzzing around their neighborhoods.
If you want to automate deliveries with class 5 robotic vans and satellite delivery vehicles this at least makes sense and serves a purpose.
Re: (Score:1)
Re: (Score:2)
Found the guy who laughed at Edison claiming you can make light without candle wax.
Common don't post AC, we want to know which luddite we're mocking.
Humans (Score:2)
Humans are too stupid, use robots instead and take them out of the equation.
Re: (Score:1)
If we eliminated the federal minimum wage then they wouldn't need to develop robots, automated drones, and self-driving cars.
Re: (Score:1)
Test your theory on a different country first.
Re: (Score:2)
Denmark and Switzerland don't have a minimum wage on a national level. But they have unions and trade associations that set one for most industries. I don't think that's better than what most of the developed world does, but it's at least not a theory.
I do believe that if we want to reduce the amount of menial work for humanity that we need to continue using technology to amplify the productivity of workers. We should have tractors instead of pulling plows by hand. We should have elevators that anyone may s
Re: (Score:2)
Re: Humans (Score:2)
We need LESS humans working, not more. Robots should be doing our work. Let robot labor subsidize and pay for humans to stay home. I mean that is how retirement plans work right? Someone else does work on your behalf, while you get paid. Well with robots it can be done ethically. If I were getting a robots paycheck, I would not mind it if the robot came up with +5 Insightful slashdot comments for me while I relax on the beach.
Re: (Score:1)
In a post-scarcity world, you just end up with oligarchs who create artificial scarcity so they can continue to profit. Kinda like how in ST:DS9 Quark sold replicated food.
So yeah, robots will be doing work, and the income goes to the owners of the robots.
Re: (Score:2)
Only 1.4% of workers are paid the federal minimum wage.
Eliminating it would make no difference.
Re: (Score:2)
Re: (Score:2)
Kind of explains why gig worker companies are so against calling people employees and why Amazon doesn't want warehouse workers to unionize.
They'll never get their satellites into orbit (Score:3)
Re: (Score:2)
I know that's a joke, but the difference is price. If Amazon is willing to pay $3,000 per delivery, then drone delivery is easy (much easier than satellite launches). The only reason it's a hard problem at all is because they're trying to make it cheaper than a person driving around.
And tbh I don't have much hope for them making flying cheaper than driving. But if they do, then cool.
Re: (Score:1)
And tbh I don't have much hope for them making flying cheaper than driving.
Flying is a lot cheaper than driving when a human doesn't have to fit inside the aircraft. The electrical costs to operate a drone are negligible.
Re:They'll never get their satellites into orbit (Score:4, Insightful)
Re: (Score:1)
Absolutely. If there was, say, a software glitch that made all the drones above a city fail, this could be an equivalent to this city being bombed.
This is probably the case of management pushing a technology hard because of how much money it could save them and not listening to people telling them that it's nowhere near ready. A lot like Uber and self-driving cars.
Re: (Score:2)
Once a drone is in the air, it fails by falling, and it it's in an urban area, it falls on something or -- worse -- on someone.
And planes are literally raining down from the sky constantly right? A failure is a failure. The risk associated with that failure can be addressed. It's quite interesting that they had a design that was supposed to land a quad with a dead propeller, sure it failure, but I've seen it work.
There was a great TED talk a while ago showing how drones could be controlled in failure scenarios. Two of the 4 motors had their props removed and yet they were still able to control the drone's position in space (though
Approve (Score:2)
Re: (Score:1)
Amazon is employing some really low IQ people. They couldn't even figure out how to fix the button without holding.
Re: (Score:2)
Yes but you'll have to replace the key caps. [amazon.com]
Really? (Score:5, Funny)
Amazon Is Still Struggling To Make Drone Deliveries Work
Amazon should talk to the Ukrainians, they seem to have precision drone delivery all figured out.
Re: (Score:2)
the Ukranians are just "servicing their customers"
Rwanda (Score:4, Informative)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
They should also look at Rwanda. They deliver blood and medical products by drone for about a decade now.
Delivering something via drone is trivial. Delivering something via *autonomous* drone is something precisely no one is doing and precisely what Amazon is trying to do.
Re: (Score:2)
Re: (Score:2)
"But I ordered *parts* for my car, not its detonation!"
The question ... (Score:2, Interesting)
... will be how does it compare to human delivery drivers for safety. The actual numbers.
Incidents aren't just magically worse because drone.
Re:The question ... (Score:4, Interesting)
It's a hard question, because the numbers aren't really going to be comparable. We have literal centuries of experience with road-based delivery systems. We've hugely evolved the design of road systems, vehicles, human training, regulations, and expectations and behavior drivers and other people using the roads.
Yes, vehicle crashes happen and are fairly common. But the types of crashes that happen, and their results, are likely to be wildly different between road-based and air-based delivery. Drones will overfly populated areas and structures all the time, while delivery trucks stick to the roads. Having drones crash and start fires -- even rarely -- is likely to trigger a different set of responses than vehicle crashes on roads.
Let's say Amazon drivers were involved in five road fatalities over the course of a month. I doubt it would be newsworthy at all. But I'm betting that if an Amazon drone crashed and burned down one house, killing just one person, it would be all over the news and people would demand that politicians Do Something. I don't even think this is irrational: currently, in our society, the expectation is that roads are kind of dangerous, but that you're basically safe from businesses accidentally burning down your house. Those expectations can change, but it will take time and we need input from more than just Amazon on whether they should change.
Re: (Score:3)
It's a hard question, because the numbers aren't really going to be comparable. We have literal centuries of experience with road-based delivery systems. We've hugely evolved the design of road systems, vehicles, human training, regulations, and expectations and behavior drivers and other people using the roads.
Yes, vehicle crashes happen and are fairly common. But the types of crashes that happen, and their results, are likely to be wildly different between road-based and air-based delivery. Drones will overfly populated areas and structures all the time, while delivery trucks stick to the roads. Having drones crash and start fires -- even rarely -- is likely to trigger a different set of responses than vehicle crashes on roads.
Let's say Amazon drivers were involved in five road fatalities over the course of a month. I doubt it would be newsworthy at all. But I'm betting that if an Amazon drone crashed and burned down one house, killing just one person, it would be all over the news and people would demand that politicians Do Something. I don't even think this is irrational: currently, in our society, the expectation is that roads are kind of dangerous, but that you're basically safe from businesses accidentally burning down your house. Those expectations can change, but it will take time and we need input from more than just Amazon on whether they should change.
This is not a "hard question". They're being held to aviation standards and the idea that they'd be held to road standards any time soon is just not going to happen for the reasons you've outlined. There's no reasonable number of kids playing in their own backyard being killed by a drone that's an acceptable sacrifice for a drone delivery lifestyle; for kids being killed in their own driveways there is to put it bluntly. The idea that seems to be the fundament of OP's argument is that if you accept one leve
Re: (Score:3)
There's no reasonable number of kids playing in their own backyard being killed by a drone that's an acceptable sacrifice for a drone delivery lifestyle; for kids being killed in their own driveways there is to put it bluntly.
Er ... why? Why is it different for in your driveway vs. your backyard?
Both involve tradeoffs of risk for "lifestyle" (which of course includes far more than just entertainment or luxury, which the term "lifestyle" tries to imply).
The idea that seems to be the fundament of OP's argument is that if you accept one level of risk you should also accept an additional risk at the same level, but that is neither obvious nor rational.
That you maybe should is both obvious and rational.
Re: (Score:2)
Additional risk. A known roll of the dice when your kids go out in front of the house does not entail that you'd also be willing to roll the same dice again when your kids go out back.
Re: (Score:2)
Why focus on the air? Walk before you fly (Score:1)
When cars break down they generally slow to a stop. When drones break down they can fly into a house. True, drone deliveries are generally meant for rural areas, but still ground-based vehicles are probably safer.
I don't know why Amazon is not focusing more on ground-based delivery bots. That seems a lower-hanging fruit, especially in rural areas where there's less to crash into.
such a pessimistic title (Score:2)
imagine this back in the early Appolo days... "NASA is still struggling with their rocket launches, another big explosion!"
Cutting edge is hard. Bleeding edge is harder. Cut 'em some slack. The payoffs are big, and the research/testing NEEDS to be done.
Honestly, we'd have more to complain about if they weren't working on this right now. There'd be so many news outlets screaming "So why isn't Amazon using drones yet?"
Re: (Score:1)
There'd be so many news outlets screaming "So why isn't Amazon using drones yet?"
I wanna know why McDonald's isn't using drones yet. It'd be like having cheeseburgers from heaven. Thanks, burger Gods!
Re: (Score:2)
Probably because most of the fast food joints are franchises. Some are corporate-owned, but not a lot.
Check it out, McDonalds is one of the largest PROPERTY owners in the world. They own the land the franchises are on, and lease it as part of the franchise. (and if you don't do well or follow their rules, they just don't renew your lease, and someone else takes over that franchise next year)
But that store ISN'T McDonalds. And instead of one jiant warehouse in the middle of some industrial park, it's a do
Re: (Score:2)
Re: (Score:1)
Their fries turn to stone in 20 minutes
Apparently the people ordering Uber Eats don't seem to mind. Every time I go to McD's there's a bunch of those gig workers milling about, waiting for their orders.
Avaition is hard (Score:2)
The Drone Delivery Cool Factor =/= THAT cool (Score:2)
Re: (Score:1)
crazy kids with drone killing guns
This is America. The kids have the regular killing type of guns.
Re: (Score:1)
Nah....an EMP gun would be more fun, and likely harder to track down for prosecution.
My Surprise (Score:2)
I'm not at all surprised by this. I fully expected drone delivery to fail, especially flying drones. I think most people did/do.
I am however very surprised at the absence of failure, so far, with food delivery drones like the Starship [starship.xyz] ones. I really expected people to be what I imagine them to be and deliver these robots to the bottom of stairwells, ponds/lakes/rivers and other calamitous endings. Or, returned to sender with steaming shit inside.
Moving fast & breaking things (Score:2)
Why (Score:2)
Why should Amazon care so much about drone deliveries? Not that many people buy drones. It's their other deliveries they should be worrying about.
Private Drone's Can't Be Flown In Silicon Valley (Score:2)
That's odd, it works for me... (Score:2)
"Amazon Is Still Struggling To Make Drone Deliveries Work" Funny I haven't had any problems getting drones delivered.