Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Drones that can haul a 20-pound load for 500 miles and land on a moving target

cnbc.com

Post reply on HN
Print view View original

31–40 of 145 posts

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#31

Post by candiodari »

a012 wrote: post #20
ShorsHammer wrote: post #15

Earlier quoted context omitted.

Jamming drones is easier than you think, the new battleground will be autonomous flying to avoid trivial radio/gps jamming. MIT has open source code for this already. https://www.youtube.com/watch?v=_qah8oIzCwk https://github.com/andybarry/flight

I'm no expert but how autonomous works if there's no positioning methods involved? (To avoid GPS/GLONASS,... Jamming)

See the paper, but TLDR: stereo video depth sensing combined with fading previous measurements (so if you detected an obstacle at 1 o'clock, 2 meters out 1 second ago, it is now very likely at 2 o'clock, 1.3 meters out. This way you can limit yourself to just detecting things 2 meters out)

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#32

Post by davnicwil »

candiodari wrote: post #25
SCAQTony wrote: post #4

Fine and dandy when flying over a desert but when the drone meets the city and an "autopilot" failure occurs over a freeway/highway/street or someone's head the consequences will be unforgivable.

I don't get that. Before the sun rises tomorrow, for the US alone, 32 Americans will make a mistake with their car that kills someone (potentially, but unlikely, themselves, 2 will kill themselves, 29 will kill someone else, one will kill 2 others, for a total of 33 deaths). At the same time, a further 2000 Americans will injure someone with a car, and another 3000 will cause severe (I believe > 1000$) property damag…

It's simply that that's not a new risk. Everyone has grown up with knowledge of that risk and accepted it as a given, therefore it is not alarming. I think this is necessary because to constantly evalute and fear ever-present risk would be debilitating.

New risks however are alarming, and do trigger the fearful 'better not try that in case the worst happens' response in a lot (the majority?) of people - this probably makes sense as a survival mechanism.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#33

Post by candiodari »

beagle3 wrote: post #5
daveguy wrote: post #2

Technology like this is why The Wall is a waste of money when it comes to deterring drug trafficking.

The technology world progresses as an arms race. I'm quite sure that when these things are used to traffic drugs, a small version of [0] will be used to stop them. [0] https://motherboard.vice.com/en_us/article/neqvag/the-us-nav...

Heh, that's cute. The issue with that is that locating UAV's that are, say 50 meters above the deck, more than 1km from your detector is essentially impossible. Without a VERY dense detection infrastructure, catching these things is not going to happen.

Furthermore even though these lasers are very cheap, by army norms, to fire, they're still 600$ or so for a single shot. So shooting down drones will be expensive. Far more expensive than the drones themselves.

Third. Defense against a laser is easy (a mirror, essentially, though you need to know the frequency to get an effective one). I must say that I don't know why you'd try to defend a drone. I'd try to hide it, since that's so easy.

And while a rules-based autopilot is easy to disrupt using jamming, a learned autopilot is essentially impossible to fool into crashing. Maybe you can divert it, I guess.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#34

Post by candiodari »

icelancer wrote: post #12
daveguy wrote: post #2

Technology like this is why The Wall is a waste of money when it comes to deterring drug trafficking.

Catapults alone are good enough technology to prove why the wall is an idiotic idea. The wall is an effigy anyway for Trump, while he may actually believe that it's a good idea, that's not how it functions in reality. It's an anchor, a symbol, a thing to point to that combats illegal immigration and is a symbol of nationalism. Now Trump might actually think it's a good idea - and I actually think he does - so that's…

Apparently, indeed: https://edition.cnn.com/2017/02/15/us/marijuana-catapult-trn...

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#35

Post by djsumdog »

icelancer wrote: post #11
cyrux004 wrote: post #3

It makes me happy to see that people are venturing out and building something that takes longer and has a higher failure rate that web apps/software.

The man who gets the most press in the tech community is trying to launch rockets to Mars and build electric supercars and networks that cost ludicrous amounts of money. Whether you think these ideas are any good or not, people are doing large things in our society. It ain't all web apps.

He did start with web apps though (PayPal).

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#36

Post by Maybestring »

sonnyblarney wrote: post #19
THE_PUN_STOPS wrote: post #18

Earlier quoted context omitted.

Read the article. The founder was an operations guy at Tesla and specifically did the numbers to justify the company based on the need of Tesla parts supply. There is room for this in the supply chain of many firms. There’s a lot of industries where waiting for something can be very expensive.

I worked in a Fortune 50 consumer electronics company, sitting right next to our procurement team. It's not my field and I'm not an expert but I know enough to be wary of a founder who makes such claims, moreover, there's enough data available to piece together what would work and not without having to accept anyone's claims on it. Most companies and their parts suppliers are in cities, wherein there are any number o…

>So the use case of 'landing on a moving sheep' is probably right up their alley

This is critical, as it's almost impossible to hold a sheep stationary while landing a drone on it.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#37

Post by hokus »

candiodari wrote: post #25
SCAQTony wrote: post #4

Fine and dandy when flying over a desert but when the drone meets the city and an "autopilot" failure occurs over a freeway/highway/street or someone's head the consequences will be unforgivable.

I don't get that. Before the sun rises tomorrow, for the US alone, 32 Americans will make a mistake with their car that kills someone (potentially, but unlikely, themselves, 2 will kill themselves, 29 will kill someone else, one will kill 2 others, for a total of 33 deaths). At the same time, a further 2000 Americans will injure someone with a car, and another 3000 will cause severe (I believe > 1000$) property damag…

About as safe as early factories. To take the discussion further off topic (sorry) It is quite easy to limit speed to the maximum allowed speed in a country. Its not hard to make a system that requires switching into city or highway mode with logging and/or remote mode detection. If we make a lot of effort we can limit speed by the amount of accidents on that location. We can further limit it by the amount of accidents the driver had in a similar situation.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#38

Post by snovv_crash »

gaius wrote: post #14
arvind3199 wrote: post #7

Earlier quoted context omitted.

Simple solution - mandate that drones are implemented with a parachute system, it's already being done on prop planes (cirrus)

A 20lb weight can kill a person from a far lower altitude than a parachute could meaningfully slow its descent. Parachutes aren’t magic.

Helicopters also have this "death zone" where an engine failure would not allow them to autorotate and have a controlled landing. And helicopters are a lot bigger and heavier.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#39

Post by mcny »

finches wrote: post #29
candiodari wrote: post #25

Earlier quoted context omitted.

I don't get that. Before the sun rises tomorrow, for the US alone, 32 Americans will make a mistake with their car that kills someone (potentially, but unlikely, themselves, 2 will kill themselves, 29 will kill someone else, one will kill 2 others, for a total of 33 deaths). At the same time, a further 2000 Americans will injure someone with a car, and another 3000 will cause severe (I believe > 1000$) property damag…

No, nobody finds that acceptable. Until you stop misrepresenting the people on the supposed "other end" of your "discussion," no one is going to engage you in a real discussion about it, which is (unfortunately) probably why you still continue to think and act this way.

I know lots of people who make rants about bad drivers but a. Fail to signal when Turing left into traffic b. Talk on the phone while driving c. Drive home from a bar "because it isn't like I'm black out drunk"

A lot of people find risks acceptable if it is them causing it.

Re: Drones that can haul a 20-pound load for 500 miles and land on a moving target

#40

Post by regularfry »

sonnyblarney wrote: post #19
THE_PUN_STOPS wrote: post #18

Earlier quoted context omitted.

Read the article. The founder was an operations guy at Tesla and specifically did the numbers to justify the company based on the need of Tesla parts supply. There is room for this in the supply chain of many firms. There’s a lot of industries where waiting for something can be very expensive.

I worked in a Fortune 50 consumer electronics company, sitting right next to our procurement team. It's not my field and I'm not an expert but I know enough to be wary of a founder who makes such claims, moreover, there's enough data available to piece together what would work and not without having to accept anyone's claims on it. Most companies and their parts suppliers are in cities, wherein there are any number o…

Car manufacturers are an interesting case, though. Despite all the efficiencies in the system, I do know of cases where boxes of replacement parts were helicoptered to the factory to avoid the contractual penalties for holding up the Just In Time system.

There's definitely a place for "cheaper than helicopter" delivery, even if it is only where things have gone so horrifically wrong that paying almost any price is worthwhile.

Post reply on HN