h a l f b a k e r y
Tastes richer, less filling.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
Android embedded OS for drones.
Phones and drones share a lot of features; accelerometers, compasses, GPS, cameras, data uplink/downlink.
AnDrone would be an open OS platform, allowing drones to be programmed via a simple API.
There would be numerous advantages.
A drone communicating via
GSM/3G/4G would be semi-autonomous given adequate cellular coverage, and combined with GPS could access mapping.
Streaming video quality would be superior; imaging could be emailed or ftp'd from a drone in flight.
There would be no possibility of conflicts for 2.4GHz channels, and drones would be "unjammable"
A mobile phone chassis provides a lot of GIPS at a low price.
Currently baking [mylodon, Dec 10 2017]
||//and drones would be "unjammable"// Why would drones be unjammable? Jamming cellphones is not difficult, shirley? (I have a small hand-held jammer, which works only over a few metres; but presumably the range can be increased.)
||Because the drone would be autonomous, and it's very, VERY
difficult to jam WCDMA over anything more than a very small
||The unit knows its start point, from GPS, and from mapping can
calculate MSA for its route. It has accelerometers which can
approximate an INS if GPS is lost. If cellular and GPS drops out, it
can either continue to execute its mission profile by dead
reckoning, or move towards the nearest cell tower, whos
positions are logged on its internal map.
||The big advantage is that the system delivers a "loiter" mode.
Simply place the drone where it can make a clear takeoff, in
advance of the mission. It could be maintained in warm standby
(RF on, GPS at low power, rotors stopped) by a solar panel or a
car battery for weeks or months - indefinitely if mains supply is
||Then, either directly commanded, cued by an SMS, or simply
based on a set of criteria being met (date, time, PIR triggered by
approaching hostile) it quickly powers up, lifts off, and does its
||As an example, such a drone could be placed above typical line
of sight, on the roof of a building, under expendable weather
protection if necessary. When commanded, it would move at
speed to its target, in total darkness and with minimal Radar
A good target would be an electrical switchyard, where it could
drop a carbon fibre mesh, tripping the system and indeed
rendering it inoperable for days. Multiple assaults from different
directions would be difficut to counter; attacking multiple targets
in this way could disable the power grid over larger areas.
||Huge resources would be needed to search for and neutralise
lurking launch points, then prevent new ones from being
established; and even then, additional devices can be almost
instantly deployed from vehicles.
||Assuming the drone is sacrificial, has a 15 minute endurance,
can sustain 30 kph in average conditions, this gives a typical ROA
of 7km, which means any target needs a protected area of Pi x
of roughly 150 km2.
||If the drone is indeed sacrificial, then the LiIon rechargeable
battery can be replaced by LTC primary cells, which have a
much higher energy density. This gives an opportunity to deploy
from longer range, or with a heavier payload.
||Proof-of-concept man-carrying drones with an 80 kg payload are
WKTE. A little gentle hacking could reap rich rewards.
||Defending against a single suicide bomber in a conventional
aircraft is problematic but possible. Defending against a dozen
unmanned drones, each approaching at speed and on a different
vector, with a 50kg payload, is seriously difficult. Engaging and "killing" such a device
is very difficult in an urban environment, since there is no pilot,
the "brain" is a very small device and easy to armour,
redundancy means that multiple lift motor assemblies must be
destroyed to bring it down, the attack can be made at high speed
and low level, (hard to engage even from a helicopter in a built up
environment), small-arms fire is ineffective given short acquire-and-engage times, and cannon- the only weapon with one-shot-one-kill potential that can be deployed (MANPADS can't currently engage such a target) - carry a huge risk of collateral damage. With 50kg of HE, packed in a jacket of flechettes, you only
need to get close. The "other" 30kg of potential payload can be
shared out between extra batteries, and ceramic matrix or non-woven aramid armour.
||All the tech to do this is available right off the shelf, right now.
||And how do you know when you've got them all? Like V-1's, the
buggers might just keep coming at you. Only one needs to get
||People or an autonomous shotgun platform would become very good at skeet shooting.
||Practical tests have shown that a sporting 12-gauge firing
standard loads of No. 4 lead shot have a less than 2% chance of
inflicting serious damage at 40m range.
||At 20m the odds improve ... but then you're inside the LBR.
||And in California, they will be copper loads, not lead, since we are attempting to reduce the volume of scavenger drones being poisoned by lead consumption.
||Will that copper lend itself to any useful electro-wizardry?