Slashdot Mirror
DRONENET: An Internet of Drones
In a series of posts on his blog, military theorist John Robb outlines what he thinks will be the next big thing — "as big as the internet," as he puts it. It's DRONENET: an internet of drones to be used as an automated delivery service. The drones themselves would require no futuristic technology. Modern quadrotor drones are available today for a few hundred dollars, and drone usage would be shared across an open, decentralized network. Robb estimates the cost for a typical delivery at about $0.25 every 10 miles, and points out that the drones would fit well alongside many ubiquitous technologies; the drone network shares obvious parallels with the internet, the drones would use GPS already-common GPS navigation, and the industry would mesh well with the open source hardware/software community. Finally, Robb talks about the standards required for building the DRONENET:
"Simple rules for drone weight, dimensions, service ceiling, and speed. Simple rules for battery swap and recharging (from battery type, dimension, etc.). Simple rules for package containers. Simple rules for the dimensions and capabilities of landing pads. ... Decentralized database and transaction system for coordinating the network. Rules for announcing a landing pad (information from GPS location and services provided) to the network. Rules for announcing a drone to the network (from altitude to speed to direction to destination). Cargo announcement to the network, weight, and routing (think: DNS routing). A simple system for allocating costs and benefits (a commercial overlay). This commercial system should handle everything from the costs of recharging a drone and/or swapping a battery to drone use."
1) Drones keeping each other informed of their vectors for distributed traffic control.
2) Additional revenue for Internet service provision to wide area near-lines-of-sight of sight to the drones current aloft. This has the added benefit of actually bootstrapping Paul Baran's original intention of packet-switching: route around the damage which, in this case, is damage to the Internet now potentiated by increasing centralization of internet infrastructure.
3) IPv6 offers the potential to finally put into place what I called "the primary discipline of network architecture" when I was designing Knight-Ridder/AT&T's multi-city videotex architecture back in the early 80s: "The terminal is merely the host computer nearest the customer." Getting rid of the client-server paradigm is key to recapturing the internet's potential.
Get in touch with David P. Reed regarding the strategic approach to take for wireless mesh networking in this new regime.
"I'd strongly encourage people today to ignore the IETF, and get focused on mobile, unlicensed wireless, highly reconfigurable and pervasive networking. Pursue overlays and co-existence, and create the next bigger "Internet" - the universal glue for networking things together. "
-- David P. Reed
Open Cobalt's synchronization architecture is a good option for an open peer-to-peer network synchronization standard currently in operation. But, as I said about the wireless mesh standard, contact David P. Reed, as this synchronization standard is based on Reed's PhD thesis, which, with minor modifications, I adopted for videotex architecture clear back in 1982 and it still has no RFC.
Seastead this.
current batteries only allow for ~~ 15 minute flights on multirotors, and this diminishes significantly as you add weight. You may be able to deliver small bags of weed or something a mile or two away
The hackers are already wringing their hands in anticipation.
Local delivery for packs of gum! Yeah!
Well I'm not criticizing it.
Just when we use the Internet analogue, imagine a DDOS with drones :)
The drones fly in the sky and when they decide to kill us all, we'll at least be expecting it.
1) incompatibility with existing aviation
2) too easy to steal a drone for all kinds of purposes
3) the terrorist implications are awesome
4) too many legal implications for accidents which are extremely
However, Fedex and UPS may be good candidates for using autonomous aircraft.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
If internet goes down then we can use RFC 1149.
Good luck on persuading the authorities in New York, London, Paris etc that you want to fill the urban airspace with drones. How will you manage 'collision detection' in its very real and physical form if this becomes popular and there are large numbers flying around? How will they talk to other air traffic (e.g. police / ambulance helicopters)? Will they be able to re-route if emergency services wish to land rapidly? How about environmental lovers concerned about the effect they'll have on local birdlife?
I guess it is manageable but a> you'll need some fine control systems to be built b> you'll have to persuade the public that lots of little buzzing machines are a good thing and not annoying and c>you're insured against the occasional fail (if you have "millions" operating, I expect some will fall out of the sky, 5kg dropping from 100m in the air onto a city crowd?)
The ability to have autonomous vehicles is immensely powerful. Flying drones is stupid and can be dismissed out of hand - but the technology to do this with the road network is already here.
Let me hack the network then net/otherwise intercept this drone delivering a 500 galaxy S iiiii.
Silence is a state of mime.
What is "GPS already-common GPS" exactly?
The system has to be able to handle delivery of more useful items than misplaced iPhones. Like groceries. A standard tote container (22"L (550mm) x 15"W (390mm) x 10"H (250mm)) is probably the minimum useful load size. There really isn't much demand for moving envelope-sized objects around any more. This is the same reason that pneumatic tube systems remain a niche product.
It's possible to scale up battery powered quadrotors to that size. But they get a bit large for urban operation.
Check out this ball drone from Japan. This can be operated safely in tight spaces.
At least this idea isn't new in trying to use alternative means then road haulage to deliver packages. In The Hague the horse and cart was RE-introduced (we are not that backward here) to deliver packages in a shopping street because it looked nice. In Amsterdam packages can also be delivered by the canal to avoid the roads. These are old techs, re-introduced. But newer versions are in Utrecht were shops have their good delivered by large trucks to a central location from which they are then distributed to the actual shop by small electric carts pulling small wagons, this to avoid large trucks blocking the roads which are one way streets mostly (no way to pass a parked truck). The bike messenger is a feature of many a large city for a long time now.
A reverse idea is in place in some crowded areas were waste is sucked into the ground and away to be collected in a central area.
These ideas are all nice BUT they lack the flexibility that the motor car has brought to our world. I can send packages by truck that would kill a bike rider, that same truck can also carry a single envelope to its destination. The only reason to use alternatives that are never as flexible, are because trucks take up a LOT of space. But you also already need the roads anyway so it is only in the most crowded areas that alternatives spring up.
Now... what is it about inner cities that makes me worry about aerial delivery by small drones. Massive buildings of wildly varying heights which already make it an adventure to cycle or even walk, let alone fly an underpowered drone carrying god knows what. As a mailman one Christmas, I delivered bottles of mercury to a dentist. Dropping out of bag, it did no harm, dropping from 100 meters...
Please correct me if I am wrong, but ambulance helicopters do NOT land between two high-rises in NY do they? One cliff is already dangerous enough I would think. What makes you think that a far smaller device, flown by cheap electronics or an operator who connection might or might not work (again, inner cities are not known for their un-interrupted radio signals, for that matter, high-rises mess with GPS too).
This idea doesn't need new tech? I know of no drone tech that allows it to operate on its own in complex 3d environments. Military drones have an operator and fly in clear skies and land and take of in wide open areas. Search youtube for remote controlled airplanes and SEE the "state of the art" tech. Especially the landings. Whooo! I want THAT going on all around me. There is a reason these people operate on remote fields with rules like "NEVER FLY TOWARDS A PERSON" (often broken but only with other members of the pack).
It is a nice idea, small drones flying around carrying packages but it just isn't practical for a long time until AI's are a LOT better, this isn't just about landing a 747 by autopilot were the auto pilot doesn'thave to think but just follow the math rules programmed into it. An AI drone needs to fly around people, unknown obstructions, unpredictable weather and all that with a cargo shape/weight behavior that is unpredictable. Box with weights rolling around inside it, good luck stablizing that AND dealing with a sudden gust of wind.
Seriously, look at youtube and the art of drones, we are still in the steam powered car era. Someday maybe but not today and not this decade.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
I use weather spark, but it would be interesting to have weather "balloons" trolleying around dynamically displaying the location of any tornado warnings.
MatterNet.us is a Silicon Valley startup who's founder Andreas Raptopoulos is one of NPR's All Tech Considered's "5 Nerds To Watch In 2013".
I've already specced out the plans for my "retrievable drone capture net", almost ready to patent it. It's going to be a profitable year!
I wouldn't even pay this for weed. I'd put money that it is also slower than USPS First Class / Priority mail. That is 2-3 days. Express mail is also probably cheaper.
Now maybe this is the cost for a mega flying truck load worth of goods rather than the lite load that I'm thinking. However California to New Jersey isn't an unreasonable distance. This is fairly typical. Where this could possibly work is beyond me.
http://www.huffingtonpost.com/2012/03/23/tacocopter-startup-delivers-tacos-by-unmanned-drone-helicopter_n_1375842.html
What if people get annoyed by the drones flying over their house, and shoot them down with catapult? Also flying over people's house could cause some privacy concern... I could mount some cameras on my drone, and fly over Girl Next Door (TM)'s house when she's having a shower.
The bigger problem is that NYC, LDN, Paris, etc, don't need this at all: we already have last mile delivery infrastructure...
Right now, 150 restaurants will deliver food directly to me. Same for local liquor stores, supermarkets, pharmacies, etc. I'm told illegal drugs work the same way. For point-to-point one-offs, messenger services fill the need. Same for large items via van service. Hell, I can even ship my kids to playdates as needed via town car.
Sheesh, I've been in the middle of a public park and ordered 10 pizzas for a birthday party. 30 minutes later, a bike delivery guys shows up with them.
Each drone would emit a radio signal . When other drones detect that signal they would slow down accordingly. The radio signal could encode location as GPS coordinate and altitude, as well as velocity.
Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
No, it ain't going to happen.
Fleets of small multicopters overhead would get annoyingly noisy, and they'll crash fairly frequently - quadcopters fall out of the sky if any component fails (motor/prop/speed controller/flight controller/battery). And in urban areas, a crash has a fairly high chance of either hitting somebody directly, or causing an accident indirectly by distracting a driver.
Electric multicopters also have very limited range (Most high-end hobbyist setups for aerial photography max out at around 15min flight time) and lifting capability. The only deliveries that would be worth attempting with drones would be illegal ones - getting drugs or guns across borders, for example...
From TFA: "Uncontrolled flight is possible over most of the US (except for cities, which are going to miss out on this, like so many other post-industrial innovations)."
Seastead this.
This whole discussion is about as ridiculous as the flying cars we were supposed to have 20 years ago.
Law enforcement could share the drones too! Just add infrared cameras, sound recording device, and you've got a perfect tool!
Next step will be equipping them with Hellfire missiles. We're just not ready for that yet.
For a few hundred dollars you'll get a quadcopter with a payload of a few hundred grams (maybe a kg+, but with drastically reduced flight time), a flight time of 10-20 minutes depending on battery and type of drone, GPS is not 100% reliable particularly in an environment with plenty of reflective vertical surfaces like buildings (multi-path signal problems), and charging high-performance batteries of the right type is non-trivial because they are prone to overheat and burn -- lithium fires are not fun and the charging has to be supervised and within a special fire-retardant bag. Avoiding narrow obstacles like power lines would be a challenge. Avoiding quickly-moving aerial objects such as other drones and birds would be even more challenging. And 10 mile range? Not a fricking chance. Not with any decent degree of reliability given variable weather conditions, and not for a drone costing "a few hundred dollars". It would have to be BIG to be all-weather. Many thousands of dollars.
None of that addresses the legal liability issues ("Your drone landed its peanut butter delivery in my swimming pool full of chocolate!" "You got chocolate in my peanut butter and on my drone!"). And operating autonomous drones outside line-of-sight and without the ability of manual intervention is not currently legal.
File this in the "give it another 10-20 years, then we'll see" category. Current technology might make it practical to reliably deliver USB flash keys between nearby (few hundred metres) buildings on a clear day without high winds. That's about it.
but think about the utility of having a container standard for smaller objects that would be
- nestable
- stackable
- designed to be manipulated/carried by robots
- has a self-describing tag or at minimum a reference in a standard form to an internet object
that you could use interoperably in a variety of storage/transport applications
Oh, and last major item delivered to my home was a christmas tree. Bought it from the Serbian guy four blocks away (discount for dec 24th, after Christmas he has to rebrand the trees as "new year trees" and sell them to Russians.) Delivered free to my living room in ten minutes by a homeless guy working as a tree hauler ($20 tip, he said he delivers 12 trees a day average, more power to him.)
Autonomous unmanned flying vehicles operating outdoors with minimal supervision? I think we'll see secure Internet voting before we see that. I think cars that are able to drive themselves will happen first, too. Just for starters, what happens when the wind speed exceeds the speed the vehicle can achieve? It would have to land safely... and then what? Not to mention avoiding power lines, staying out of restricted airspace, making the system secure, batteries dying at inopportune times, communications issues, liability for damage caused, noise complaints, mechanical issues, ...
What a fool believes, he sees, no wise man has the power to reason away.
Than theorist John Robb. I am certain he is right his time table is just a bit off. We are slow to accept anything we always have to be dragged kicking and screaming to the money.
I'm pretty sure some guy have already failed to start a drone delivery business.
A mesh network of transport vehicles would be a very efficient way of moving items of varying sizes.
The courier system could be augmented by air and water based drones. Small packages make up a significant amount of the total and the last mile problem applies here too.
A potential solution is bulk heavy transport to hubs as happens now and then drones for the final hop rather than using road. Of course swapping loads between drones would also be useful.
Properly implemented this could use less fuel and be more flexible than current courier services.
The problem of loss of cargo would be managed through better redundency design in the drone hardware/software. Noise and danger to people underneath could be managed by defining corridors into the drone map, for instance traveling above a rail line should be relatively safe as should be travelling on water.
Some comments highlight legitimate criticisms to this ideas adoption everywhere, in particular in cities.
However combined with other distribution methods, this could become part of a wider open delivery system where an item could use various means of transportation to be ultimately delivered to its destination.
I also like the idea of having the drones double as mesh Internet providers.
How are the drones going to not run into each other ? How are drones from the "older" generation: the ones not on DRONES.net, going to avoid collisions with drones that aren't on the network and that aren't able to detect other drones?
The day a delivery packet fells on a baby is the day that technology is outlawed nation-wide or continent-wide...
I'm sure teenagers will have a great time downing these things. Stone plus string makes a rotor-tangler. Just use atlatl or traditional cartoon dennis-the-menace slingshot. With luck, you get someone's delivery. At the very least you get a mangled drone to strip for ebayables.
this is just to help fund hte usa's latest military wack a mole drone bullshit.
waste a my money cash , time , effort and most importantly some ones life.
go try and make the world a better safer place first before you come back spouting you shit...omg and they think OH JOY can i BUY ONE and help fund world destruction
LOL
seriously go away get fucked and stop war mongering by another name.
https://torrentfreak.com/worlds-first-flying-file-sharing-drones-in-action-120320/
and the idea is genius. This could never happen without some company pioneering the idea in the private industry. Even if you had the drones flying over major highways, people would get freaked out at flying packages without some kind of massive marketing campaign.
Unfortunately there are only a few countries where the drones of today may be economical.
Look at Ecuador for example, they have slow infrastructure and insane mountains everywhere. Drones would connect this relatively small country very well. They have small landing fields everywhere and possible delivery needs of all sorts.
~ Best man at your service.
Reading the article the first thing that comes to mind is: Packet loss and bandwidth.
I think we will see the same problems as we have on the internet where some deliveries will just take forever or will never happen at all.
How about the reliability (a.k.a. packet loss). What happens if a drone malfunctions? What happens if the product / item that is transported gets lost?
For such a delivery service where I cannot simply resend the item (like I can do with most information on the internet) I would want some level of guaranteed delivery.
Don't get me wrong, I like the idea, but to make this work we/he needs to put together a set of rules/policies/design features to make it work reliable and with a (near) 100% failsafe mechanism
Nice for densely populated places, but for more suburban places those services don't really exist, except for pizza delivery. And it's possible that those also aren't necessarily delivered in the most cost effective fashion. You could purchases several of the super high end multirotor copters for the salary of one bike messenger, let alone anyone using cars to deliver. And it could, in theory, deliver far more quickly than a person possibly could.
I think the big problem is weight. We need copters that can easily carry 20-50lbs. These days the upper end of the requirements is an expensive video DSLR with a heavy lens and articulating mount, maybe 5lbs?
Most of these are simple engineering challenges we've already solved with airplanes already (transponders, etc). I think mostly it would just be a matter of the FAA finally revising the rules for sUAS. Maybe any company operating these drones has to electronically send flight paths in real time to the local air traffic controller? There are also ways to make sUAS that are safe for flying around birds and people.
Who's going to accept the liability? There roughly a billion dollars of liability for each aircraft. People have even been sued just for advertising they're rich by having a ($20k) airplane. The FAA has been dragging their feet for 5 years on this and have no motivation to createmrules. Given the percentage of congress which is full of lawyers, I don't blame them. The Bly way we'll have sense and avoid is trough the DoD using Title 10 (fuck you, we have guns) authority.
This will bring on a new breed of criminals. Drone robbers. Bring a drone down as its taking off or landingt and steal the cargo. Or better yet - drone crashes due to mechanical failure. Contents are then stolen. Maybe even the drone itself is stolen.
Yes, I can see that working. But who's going to pay for the emergency services helicopters' radio equipment to make sure drones don't fly into them? (apologies if this is trivial, I don't know what equipment police/ambulance helicopters carry)
Go ahead, put all those quad-rotor clay pidgeons up there. I'm a hillbilly with a 12 guage, what could go wrong???
My karma is bad. Don't get too close!!!
I see a potential platform war as people try to standardize, vying for a de facto standard.
Tacocopter vs Burrito Bomber. Who will win?
I wish they will now see the difference of building something against the people, compared with Internet...
Below are some thoughts to add to this fascinating brainstorming session. I hope that some of them will turn out to be useful in generating further discussion.
General concepts.
Naming
Please can we have some better name suggestions for this system? Drones already have a highly nefarious reputation and dronet sounds like some malevolent black ops military program. How about delivery bots or auto hoppers? Please anything by drones. Some Westerners might not realise how much resentment that the term drone stirs up in many parts of the world, and this could be an unnecessary obstacle to adoption in places where the service would be the most useful.
Uses
Unlike other posters, I do not feel that fast food deliveries are the killer app for these devices. High density down town areas are already well served by bike couriers and delivery services. UAVs flying around central business districts will be dangerous, aggravating and a nuisance. Think of all those wires in which UAVs could become snarled in and the lack of open horizontal spaces for landing.
This concept is much better suited to less developed areas that are not currently served by dense transportation networks. I predict that the first early adopters of this network will be the growing number of resilient communities who want to be able to leverage their surpluses with other similar communities. Locations that are not well served by sealed roads and where topography and vegetation causes obstacles to conventional transportation methods could benefit greatly from delivery bots. Lets face it not everywhere is suited to cars and trucks.
If I live in a big city then getting delivery for just about anything is already very easy. If I choose to live away from a major population centre then things are much more difficult. Centralised distribution networks still reach me but add to their overall inefficiency the further away that I am located. A distributed network can overcome such problems. I can already think of many permaculture smallholdings with so much surplus production that a landing pad would be ideal for their needs. They have plenty of high value items to offer and can request useful manufactured items in return. Most permaculturalists I know have large excess stores of value added items. These include exotic fruits, herbs and spices, baked goods, preserved, smoked and fermented foods. In addition they have large quantities of seeds, seedlings and even surplus living creatures such as small mammals and insects that are beneficial to a flourishing well stacked ecosystem. Then of course, there are manufactured items, tools, books, spare parts, craft items, etc.
A Gift Economy Based on Abundance Rather Than Scarcity
These might seem like strange items to exchange over a network of autonomous delivery bots until I explain that I do not see this as a paid subscriber service, but as a system of ratios and reputations. If this network is going to be distributed in the true sense of the word then we cannot have large ISP type corporations creating artificial bottlenecks for their own personal profit. Instead I hope to see a network where landing pads and their associated technology are so cheap and ubiquitous that anybody can install one.
The next stage would be to offer items for collection that would encourage the UAVs to actually come and visit. Setting up a landing pad and a fee based system where individuals need to pay to have things delivered will be far more complicated than a network where every node lists all the goodies that are available to be collected. This is a system based on surplus rather than scarcity, and where the complicated disadvantages of modern finance are unnecessary. If I want to encourage a delivery bot to come to my pad then I list as many items as possible that the UAV can collect and delivery onto another node. First I am going to list all those books, magazines and DVDs that are taking up space in my cupboards. Next I am going to start scouring my home for useful things
b> you'll have to persuade the public that lots of little buzzing machines are a good thing and not annoying and
same with cars
c>you're insured against the occasional fail (if you have "millions" operating, I expect some will fall out of the sky, 5kg dropping from 100m in the air onto a city crowd?)
same with cars again, a car weighting a ton will sometimes crush into something eventually.