Musing over the idea that drones (and by this I mean the little ones, usually small quadcopters, not multi-million pound aerial weapons flown by remote) will have a major role to play in modern urban societies, I concluded that it would be both fun and instructive to work through just what their role might be.
Both Google and Amazon are experimenting with drone delivery systems and I can see the appeal of this immediately. Instead of having to wait a whole day for gratification, the lengthy gap between ordering online and when our newly purchased parcel arrives, we can have our gratification almost immediately if we can organise a drone to deliver the purchase instead. We see that thing online that will make our lives either complete or a bit less incomplete, we buy the thing, and a drone delivers it to our door a mere hour later.
Lets just take it as read that this will happen in any case, because if for no other reason, there are a lot of pizzas that need delivering every day, and this would take a lot of maniacs on scooters off our roads, and thats the thin end of the wedge. With the growth of online shopping I have heard a TFL* representative say that 30% of Londons traffic is delivery vehicles. Even if we took 50% of the vans off the road, we would reduce congestion a lot, reduce emissions a lot as most vans are diesel powered, and the streets would be quieter and safer. Apart from the hum of drones that is. Perhaps the pigeons would disappear too, perhaps there would be too much aerial traffic for them to feel comfortable, one can hope.
But there are a lot of problems to be solved and barriers to that future.
Legality
They’re illegal, and cannot be flown near to people, which is a bit of a problem when you want them to get close to people to deliver goods to them. I think this will go away presently as the software systems running on the drones enables them to be more or less autonomous and able to avoid crashing into things or people. If cars can be considered safe as driverless objects, then drones shouldn’t present much of a challenge, being much smaller and lighter, and posing much less risk to human life. Lets assume that that challenge is surmountable.
Location
Drones don’t currently have much of a range from the signal that controls them, which means that if you are a kilometer away from their controller they aren’t much use. I think that this can be dealt with by allowing the drone to control itself and by having a distributed network of guidance, like cellular telephone masts, that provide locations to the drone as it comes close to the mast. We use these masts to locate ourselves with smartphones, so why not drones too? To get it to deliver to our houses we just need a way to broadcast a signal to it that it can recognise, perhaps like the one created by our WiFi routers?
Distance
Being battery powered the current quadcopter drone designs are limited in terms of the distance they can travel and the loads they can carry. Battery technology is getting better, so distance will grow over time. The location masts or beacons used above to tell them their location could also provide charging points, so a tired worn-out drone could stop off for a quick gulp of electrons on the way home after delivering your pizza, book, fresh coffee,..whatever. To take on heavier loads drones could cooperate. This video by ETH shows a group of drones constructing a rope bridge, and this one shows another team creating a structure using bricks. The relatively straightforward task of delivering a parcel looks rather easy in comparison.
Identity
One of the major problems with drones is privacy. People don’t like the idea of a machine equipped with a powerful camera flying over their heads on a daily basis. This seems a bit Luddish to me, after all, in our cities we are surrounded by cameras in the hands of everyone we pass as well as those on the streets and buildings. But lets address the problem anyway. Imagine a scenario where the drone is autonomous, and not under control by any external agent, as it winds its way from depot to you. It doesn’t even need to go to your house, if you are in the park having a coffee, it could deliver the pizza directly to you. What the drone needs is autonomy, and a way of getting an anonymous set of directions to you. It need never know who you are, or where you live, and even better, it need not know what it has delivered to you. This will help to avoid the problem of Amazon and Google knowing everything you ever bought so that they can try and sell you a duplicate of everything you own. (why don’t they try and sell you something you haven’t bought?)
Delivery
Lets take a scenario where you order a pizza and its awaiting delivery at the ‘restaurant’. A signal is sent out that a delivery needs to be made, and the nearest drone accepts the job in the same way that a Uber car would. The shortest distance to pick up the pizza would offer the cheapest transaction cost. The drone collects the pizza, and is given an electronic token at the same time. This was created by you when you ordered the pizza. Half the token goes to you, and half to the drone. You broadcast the token from your location and the token is passed from one point on the network to the next, every time the token is passed on it gets a bit added by every node on the network. The network propagates the token from one node to the next indiscriminately. This enables the drone to follow the trail back to you by seeking a broadcast token that is shorter than the one it has picked up from the network and which matches the other half of the token it already has. It will find its way to you without knowing who you are, or where you are.
Esch bubble represents a location beacon, such as a wi-fi router., the routers broadcast the destination to the drone, and the drone follows this to you, wherever you are. The box at the top left is the warehouse sending in the message, the box on the right is you waiting for the delivery. Drones already in the network pick up the message signal, follow it to the warehouse, pick up the package and deliver it to you.
Security
In the same way that BitCoin has developed a security system that is distributed, and every bitcoin node knows how many bitcoins there are, and who owns them, without being controlled by a central source, drones could carry out the physical transactions managed by a similar system to the electronic transaction. In a nice parallel where BitCoins enable Payer A to use currency B to pay C, the drone can carry the package from C back to A using the network B. Read this article on blockchains and BitCoins and you will see what I mean. This method would prevent anyone knowing which drone was carrying which package, and who it was intended for. The only way someone could steal your pizza would be to follow you home and steal it from the drone as it delivered it to you. Of course there will always be people who will snare a drone for whatever it happens to be carrying, but at least they won’t be able to steal on demand.
Physical Implications
A drone needs somewhere to land a drop off its parcel. It needs a flat surface to land, and if the person for whom the delivery is intended isn’t there, it needs an electronically linked drop box where it can leave your parcel. It could lock the box with its half of the electronic delivery token, and you can unlock it with your matching half when you get home from work. But the box needs to be big enough to accept your pizza, post, packages and needs to be somewhere that the drone can get at but where other people cannot. For apartment buildings this would ideally be the roof, where a landing platform and a set of drop boxes could be located without too much difficulty in many flat-roofed apartment buildings.
Perhaps one day drones will be able to post letters through your letterbox, if you still get any, any that you actually want to read that is.
*Transport for London
rory bergin's blog 