Blue-sky thinking on drones

When I worked at Storyful my colleagues often slagged me over my apparent obsession with a few things: drones and drone technology, satellite imagery (and other possible satellite applications) and other automated newsgathering systems (my own startup is one such system).

But pretty much every brainstorming session started with the first one: drones. Of course the context of those meetings was usually their future application for the news industry — but my thinking on the issue was far broader.

For me this thinking was always brought to the fore by natural disasters. I watched many, many natural disasters vicariously through the lens of witnesses to those disasters — watching many thousands of hours of eyewitness content from hurricanes, tornadoes and earthquakes — all too often content from the poorest parts of the world.

When watching these events unfold I would often wonder at how these were often actually anticipated — we did not necessarily know the precise time and location of the event, but we often knew it was coming. This leads you to think about the future, and how the combination of pre-preparedness and technology could perhaps give us a sense of how these disasters could be handled in the future.

Following recent events in Nepal, I wanted to write down some of this thinking. There are practical limits to some of what I write, but I believe these limits to be surmountable — and my time horizon is the years 2020–2030. My other proviso is that I have zero expertise in aid or the provision of emergency aid in natural disaster situations. This is merely a thought exercise.

There are a few different types of natural disaster in poorer countries (and I’m focussing on poorer countries primarily because more advanced economies are better able to handle post-event aid).

  1. Anticipated natural disasters affecting high-risk countries.
  2. Anticipated natural disasters affecting low-risk countries.
  3. Unanticipated natural disasters affecting high-risk countries.
  4. Unanticipated natural disasters affecting low-risk countries.

In any of these 4 situations different preparing might happen. There might pre-built operations in high-risk countries, with low-risk countries being able to avail of assets that can be located there quickly.

So in these circumstances, what role could drones play in the future?

Using modular systems one could imagine something akin to a shipping container. This TEU might contain let’s say 50 drones, and there might be many dozens of TEUs. Drones would be segmented by role (and configured for these roles which may require longer/shorter flights, different payloads etc) which might include:

  1. Survey drones. These would be the drones deployed as early as possible — either from containers flown to the location, or from pre-built units in high-risk countries. Once an event has occurred these drones would come with pre-built censors — data collection, 3D mapping, heat signature counting — in essence an automated survey of a radius. And then multiple radii within a pre-specified survey area, or a survey area that evolves based on what the drones find. These drones would be smart enough to initially assess damage and then feed back estimated needs for food, water, and supplies for reconstruction. A typical task might be survey drones following all supply routes/roads to assess damage, taking video in realtime and that being recorded by drone control rooms. Conceivably these drones could also listen for survivors.
  2. Delivery drones. These drones are tasked with delivery of vital supplies to locations designated either by survey drones already at the location, or based on human surveys/live satellite imagery gathered in the first 24 hours post disaster. Supply packs would be modular and pre built — medicines, food and water etc. Whatever aid agencies deem appropriate. Of course there are weight and range considerations, but I believe with time these will be addressed. These drones would operate in coordination and 24 hours a day, 7 days a week. Many hundreds of drones could work in tandem, with automated re-charging (solar + batteries would refuel batteries that drones depend on) and discrete tasking orders assigned by human controllers.
  3. Communication & relay drones. In order to minimise anxiety in disaster zones, drones would act as relays for important information — such as displaying information on when the next delivery drone is due to arrive (“Next food drone arrives in 90 seconds”). These drones could also act as remote networks, allowing basic internet and cellular connectivity for affected survivors. These would be rotated automatically within the power/flying constraints of the drone. Relay drones would act as conduits in the network, but might also have automated management functions to ensure optimum flow.
  4. Infrastructure drones. These drones would be designed to land in areas where there are immediate infrastructure deficits — drones specifically carrying batteries, solar cells, satellite comm equipment etc. These would only arrive after an adequate supply drone chain has been established. But in theory these drones could carry anything (within weight restrictions), depending on the needs of affected victims.

Of course the drone “home” might not just be a pre-built TEU. It could be a pre-positioned blimp which stays aloft for long durations. It could also be dedicated buildings designed for just such events. Or it could be that within certain buildings a space is set aside for drone storage/launching.

Once the initial needs of a population are being met in the first 72 hours of a disaster, normal supplies may kick in depending on how well traditional methods are evolving. But in theory, once there are adequate supplies of modular packets of food/water etc (and that’s a big problem in itself), one could imagine that such a self-contained and self-powered fleet of drones could operate for weeks at a time, with little human interference.

There would be interesting software to be built in all of the scenarios outlined, including:

  1. Task management systems. ATC systems for thousands of drones.
  2. Image recognition & 3D survey systems (and rapid integration into bigger data).
  3. Learning systems which adapt to changing circumstances based on human behaviour + evolution of natural disaster, affected human assets.
  4. Supply chain systems — how needs are assessed automagically by machines and orders placed for replacement commodities, material, equipment etc. (And in turn supplies are automagically shipped to required destinations without human intervention).

I find thought exercises such as this useful in pretty much every industry and every facet of technology. Some of it may come to pass, some of it may not. But it’s often enough for something to be conceivable to make it at least possible. And no doubt many other people are having the same ideas, and starting to build these kinds of systems.

In the future, will machines play huge roles in delivering aid to humans affected by natural disasters? Definitely. Will it be as outlined above? Perhaps.

Gavin Sheridan is founder of Vizlegal, a global platform for law (always under construction!), and former Director of Innovation at Storyful. He has a cat called Puds.

Founder/CEO @Vizlegal | FOI, journalism, law, data | Former Innovation Dir @Storyful | Dublin, Ireland

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store