Imagining a Drone Delivery Service

This week, the BBC published a piece about the Hermes delivery service, which is beginning to experiment with drones. while I was quoted in the final article, here are some more extensive comments on what drone delivery means for us.

Hermes are overcoming many of the technological challenges that face autonomous vehicles today. Their successful testing of the robots over the last year shows a proof of principle and practice. So, once the technological challenge has been solved, the cultural shift is next. We have moved into a one-click era, where we expect to consume everything whenever we desire it. We don’t want to wait at all. That might be fine in some circumstances, but imagine if our weekly grocery shop became an itemized affair. Individual robots delivering each item, as and when you desire them. This is already happening with Amazon deliveries, but would be on a much bigger scale.

So, unless we modify behavioural expectations to avoid this highly atomized approach, we could end up with millions of robots in operation, which means a lot more logistical challenges. At the moment, in big cities, it may be hard to walk down the street due to the amount of people. In time, it could be the amount of robots we are trying to dodge and I’m not sure that would be a better world.

However, drone delivery is a logical extension of our present day consumer culture. Since the dawn of shopping tv channels like QVC and then Amazon, it is clear that the buyers’ desire is to have their products as soon as possible. The last mile delivery challenge in built up areas especially is huge, but self driving robots does sound like the best solution.

There are different challenges for ground or air based systems. For instance, in the sky, we still have to figure out what a drone highway looks like and figure out some sophisticated collision avoidance technology. On the ground, at the speeds Hermes are going,

Redundancy in the work place is a huge likelihood, so we need a radical redeployment strategy to ensure massive of service staff are not out of work. We have seen this coming with the industrial age and we’ve mostly managed to re-skill, but this scale would be unprecedented. It’s not just delivery of items, but services too. And in a service economy, this can be devastating.

It’s not a bad idea to make robots pay taxes in the same way that we do, perhaps to support those people who find themselves made redundant as a result of it.

We also need to be aware of the problems that might arise from systems being hackable. Once we lose control of the platform and hand over control to a technological system, then we make it vulnerable to this kind of cyber attack. This may be the single most important tech challenge facing autonomous vehicles and, but the number of incidents may still be less than those that occur as a result of human error. However, this may not be much comfort to people.

What are the advantages of such autonomous delivery methods?

The obvious answer is that it brings efficiencies into the labour force. Once a robot delivers, there is no need for humans and that means more than just cost cutting on salaries. It means also not having to worry about workers’ rights, holidays, and so on. But there is much more to it than this. One of the biggest advantageous actually reflects a broader principle that underpins economics, which has to do with predictability. Remove the human and systems become more predicable, easier to refine and improve. The biggest savings are achieved by this enhanced predictability.

Aerial versus Ground Based Systems

Large scale aerial delivery systems are likely to become promient after ground based services are established, but they will each have a different role within the system. So, aerial robots might serve the last 10 miles, while ground vehicles will cover the final mile.

It’s important to see these different robots as having different roles to play within the system. There will be no single robot solution, but a whole network of different sized robots with different capacities. Some will need to travel big distances with large pay loads, while others will need to be agile and able to move through small areas.

There is a lot of push back against aerial drones because of safety. This problem is likely to be overcome, but that doesn’t mean we want to have the space around our heads constantly buzzing. Electric ground vehicles will be more desirable for this reason alone. They will be more able to blend in and will most likely be multi-functional systems. While delivery might be their primary function for us, their data gathering or sensing capacities — gained through the act of delivery — might have the biggest value commercially.

In this respect, our consumption of the product through the delivery of something, actually serves a greater commercial goal. It’s a bit like how social media works as an economic proposition. We share over and over and get something personally valuable from that, but these acts serve a bigger model driven by our data.

Getting the job done

At the moment, what’s lacking is a comprehensive design solution to a world with multiple kinds of robot delivery systems. We haven’t figure out how to connect up all these services, or figured out what would be the best way to optimize delivery. It’s likely that, like the emergency services, the optimal systems would be publicly owned. To have multiple robots with multiple capacities is less desirable than to have simply the single best system out there.

Moreover, to have multiple robotic systems all configured to work around each other so that there are no mishaps would be hard in a completely private, commercial model. It isn’t like cars, where we can choose the one we want to drive. These vehicles will have to work together, as part of a system. At the very least, we need to ensure they have common intelligence.