h a l f b a k e r yThunk.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Drones have the promise of delivering products to
people's
doorstep. There are issues with this, but assuming this
finds popularity, a simple way to avoid mid air collision
might be to have all drones that are flying at the same
altitude fly in the same direction.
So any given area, your
map is broken into degrees on
the
compass. Where you're going would determine how high
you fly. You'd have 36 directions each having its
designated
altitude.
North = 500 feet.
North + 10 degrees = 510 feet
North + 20 degrees = 520 feet etc.
This would afford a great deal of separation between
drones flying in opposite directions and close proximity
passes only when they are flying almost in the same
direction.
A standard speed would be necessary to keep drones on
the
same path from overtaking and crashing into each other.
Something like this?
Anti_20Altitude_20Fixation
[ixnaum, Aug 09 2016]
Possibly the Skyline scene [bungston] mentioned?
https://www.youtube...watch?v=TNTD1j8YRuM
Dumbest way to deliver a nuclear warhead I've ever seen. Also cringeworthy filmmaking/acting in general. [notexactly, Aug 18 2016]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
sounds like Anti Altitude Fixation but for drones (link) ... makes even more sense with drones [+] |
|
| |
"Airways" for real aircraft are Baked & WKTE for a very, very long time.... |
|
| |
I was thinking about 3d space for planes on watching a clip from a movie called Skyline, in which a (legion?) of drones attack the alien mothership. The drones approached in a horizontal line along a plane like a bunch of Greek hoplites. I suppose that makes it easier to show but did not make a lick of sense - an alien defender would get a chance to attack multiple drones as they approach, and closely spaced the drones were at risk from debris due to adjacent fights. |
|
| |
If I were commanding a drone attack I think they should come in from all points of the sphere around the floating target. |
|
| |
Ok doc here is your challenge. Your flight corridor is all fine and good for drones tooling along the level. But I suspect most civilian delivery drone courses will be an arc of some sort. Bust out that calculus and show how your scheme can deal with overlapping and possibly interecting arcs of various sizes. |
|
| |
I think this is the wrong kind of system. It would be
more better to ensure that drones are aware of
eachother, and then give them some basic flock-
behaviour algorithms. This would allow for more
flexibility and greater efficiency. |
|
| |
//"Airways" for real aircraft are Baked & WKTE
for a very, very long time....// |
|
| |
This isn't that. Airways are simply designated paths in
the sky, exactly like roads only you can't see them. |
|
| |
This is a very cheap, very easy to apply dynamic air
corridor designation concept that can be used
anywhere on the planet using only an altimeter
and a compass. No ground control, no collision
avoidance radar and most importantly, no designated
air corridor maps you have to follow. This can be
used anyplace, over the Arctic Circle or the Gobi
Desert, the paths are the same, if you're going North
+30 degrees, you fly at 530 feet. |
|
| |
Only flaw with this is if you fly over the North Pole
you go from flying north to flying south and have to
change your altitude. |
|
| |
The problem of drones flying across the North or
South Pole and having to change their altitude would
almost never arise so not a real problem. |
|
| |
As far as the arc flight paths, they'd have to fly a
straight line. They'd have to travel vertically to and
from flight altitude to achieve this. You would have a
brief crossing of paths if you flew through other
corridors getting to yours so it's not perfect. |
|
| |
Ok, woke up with the solution to the ascending /
descending collision problem. |
|
| |
When a drone takes off, it emits a high frequency
sound too high for human ears but readable by other
drones. As it ascends, it registers its height by the
frequency of the sound so approaching drones can
hear that there's a drone encroaching on its flight
path. |
|
| |
If the approaching drone hears this, it emits a
specific multi frequency beep that the other drone
can hear, like a car sounding its horn. The beep
consists of a warning frequency and the approaching
drones altitude. The drone that's taking off or landing
hears this and moves above or below the approaching
drone's path. |
|
| |
Neat side effect of this is a drone taking off would
make an ascending sound like a 1950s sci-fi flying
saucer taking off and a descending tone as it landed. |
|
| |
Although it would be beyond human hearing. |
|
| |
Could drive the dogs nuts though. |
|
| |
Howabout just a coordinated system in which they already
know eachother's positioning? Automated drone air traffic
control central? |
|
| |
// a coordinated system in which they already know
eachother's positioning? Automated drone air traffic control
central?// |
|
| |
No no no. Central control will be a disaster. Bees, fish and
birds have this all worked out, and fly in dense - sometimes
interpenetrating - swarms yet never collide. And they have no
central control. |
|
| |
It is incredibly easy to simulate bird flocks or fish shoals. All
that is needed is a handful of internal rules which each
individual follows, and an ability to sense nearby flock/shoal
members. |
|
| |
Flocking rules are extremely robust: even if one bird goes
nuts, the others avoid colliding with it by following the
flocking rules. And each individual can go wherever they
want by the most efficient collision-free route at the time.
Likewise, even a non-compliant drone blasting through a
crowd of flocking drones will not cause a collision. |
|
| |
//Howabout just a coordinated system in which they
already know eachother's positioning? Automated
drone air traffic control central?// |
|
| |
That would require network connectivity and central
control of all the flying units. This seeks to avoid
that. Anybody can fly anyplace around the world with
this system, no central control or radio
communication necessary. |
|
| |
//No no no. Central control will be a disaster.
Bees, fish and birds have this all worked out, and fly
in dense - sometimes interpenetrating - swarms yet
never collide. And they have no central control.// |
|
| |
Yes, as you say, central control is to be avoided, and
of course, first we see how father nature has
achieved the job. Problem is, swarming in nature is
all about flying in the same direction so not much to
go on when trying to figure out how to have
thousands of things flying in all directions. |
|
| |
Now of course, one may ask, will there ever be
thousands of things flying in all directions? Dunno,
but if there were, this would be the simplest way to
keep them from colliding. |
|
| |
//swarming in nature is all about flying in the
same direction so not much to go on when trying
to figure out how to have thousands of things
flying in all directions.// |
|
| |
Flocking algorithms still allow for each member to
have its own agenda - two flocks of birds can
interpenetrate, for instance. In fact the same
rules will allow every member of the flock (which
would no longer look like a flock) to have a
different preferred direction. The flock rules only
serve to avoid collisions. |
|
| |
Either system might work, but a flocking system
will continue to work (and remain collision-free)
even if there are rogue elements that don't follow
the rules. |
|
| |
Problem is, you're asking a lot of the sensory
mechanisms and their reaction times. |
|
| |
I interpreted this system to be integral to the drones, not something imposed on a dumb machine by a pilot. The drone knows path and would choose altitude accordingly. Audio communication with nearby drones would also be automatic and would work if the drones were not too crazy fast. Radio instead of audio would work for fast drones; still autonomous. |
|
| |
The problem would be rouge drones which declined to communicate with the commoners, and were also a very undroneish shade of red. |
|
| |
//you're asking a lot of the sensory mechanisms
and their reaction times.// |
|
| |
Not so much. The sensors need only detect the
distance to the nearest drone (or bird! This
system will avoid birds too!) in each of six
directions (up, down, left etc) - effectively a
parking-radar, and surely easy enough to
implement. The reaction time of the electronics
will be negligible; the reaction time of the drone
will be determined by its mass etc, but you can
choose the range at which it decides to avoid
another drone. This also means that big, slow
drones can take avoiding action early, whilst
smaller nimbler drones can be a bit more efficient
by only dodging when it's really necessary. |
|
| |
I'm sure there are other solutions, this one is just the
simplest, and I also left out that I think drones need
to be hardened against RF interference. I'm assuming
when the stuff hits the fan any RF control of these
things goes out the window with jamming. |
|
| |
Both scenarios need to be tested for efficiency against one another in the 3Dunder dome! |
|
| |
Two plans enter... one plan leaves. |
|
| |
Just thought of another big advantage of this system.
No need for standardization. Anybody can make a
drone that won't crash into others as long as it has a
simple altimeter and compass and adheres to this
flight rule. No communication between
manufacturers of the drones is necessary at all. No
updates would ever be necessary either. There's
nothing to update. |
|
| |
While this would successfully achieve its stated objective,
it would do so at the expense of most of the usefulness of
drones. For tasks like mapping and urban tree inspection,
drones need to be able to fly in specific directions at
specific altitudes, combinations which likely don't
coincide with your rule. |
|
| |
The guy who runs the RC Model Reviews YouTube channel
is working (slowly—the government and the radio control
association are obstructing him, apparently) on a passive
radar sense-and-avoid system. I think that would work
well with [MB]'s suggestion of bird-style collision
avoidance. |
|
| |
//over the Arctic Circle or the Gobi Desert, the paths are the
same// |
|
| |
You may have some trouble crossing the Himalayas at 530 feet.
Or Manhattan, for that matter. |
|
| |
The heights would be customized to the area
obviously. The relative height to direction ratio is
what would be a standard. |
|
| |
This is one I kind of wish I had patented because
I'm
predicting that this is how it'll be done when we
get
drone delivery for everything from mail to fast
food. Right now the only thing really stopping that
from happening is people assuming that hundreds
of drones flying every which way over a city at any
time would
need a complicated, high tech, failure prone
traffic management system which this makes
unnecessary. |
|
| |
And think about this system compared to what we
do now. A 2,000 pound vehicle being driven by a
human to deliver a hamburger? We actually do
this. Absurd. |
|
| |