Aardvark DailyNew Zealand's longest-running online daily news and commentary publication, now in its 25th year. The opinion pieces presented here are not purported to be fact but reasonable effort is made to ensure accuracy.
Content copyright © 1995 - 2019 to Bruce Simpson (aka Aardvark), the logo was kindly created for Aardvark Daily by the folks at aardvark.co.uk
Please visit the sponsor!
As I usually do on a Wednesday morning, today I took part in a global panel discussion of things related to drones.
The panel often consists of figureheads within the industry and those who are involved in the regulatory and technological evolution of this rapidly changing area.
One of the key discussion topics this morning was the issue of Unmanned Traffic Management (UTM) and how it will be best achieved. It was an interesting and lively debate.
I took the position that the future of managing millions of drones in our skies would be something that was ideally suited to autonomous craft with a high level of artificial intelligence. The reason for taking this stance was that I frequently marvel at how much of our own hi-tech is actually an analog of the stuff that nature and evolution have developed over countless millennia.
Even the aircraft in which we fly are modeled very much on the things we've learned from nature and even though we have enhanced and extended those principles, the underlying basics were not our own invention at all.
My rationale for suggesting that UTM is best left to autonomy and AI was based on the way that humans manage to negotiate potentially treacherous environments, such as a busy city sidewalk -- or when driving on a fast-moving motorway.
Although there are points of control, such as traffic lights and "cross now" signals, the actual movement of each individual person or vehicle is left up to the "intelligence" of the person in control and most of the time such systems work very well.
Another comparison can be made with birds...
Despite the fact that in some areas there are extremely high concentrations of birds (such as around tall oak trees on an autumn evening when massive flocks of sparrows and starlings can be seen moving in perfect harmony) we see very few "mid-air collisions" between birds in these areas of congested airspace.
Despite the density of birds, the relatively high speed and the fact that flocks may, from time to time, share airspace, they all manage to get from point A to point B in complete safety. There is no "controller" on the ground coordinating and controlling their movements -- onboard "intelligence" seems to do the job very nicely.
So let's look at drones.
Unlike the situation when birds are roosting of an evening, the density of drone traffic in the skies above our heads is incredibly low and will remain so for quite some time to come.
So, surely if thousands of small birds can flock through a very small amount of space with safety, even a hundred drones per cubic kilometre of airspace should be able to find their way from source to destination without any issues -- providing sufficient AI is onboard to provide them with navigation and avoidance cues.
Why create a massive, complex, single point of failure by building a traditional air-traffic control system in the form of the currently proposed UTMs?
Surely it would be better to distribute the task amongst all the drones themselves, providing them with mesh networking and sufficient AI to navigate around obstacles and potential threats. If one drone's AI or comms fails then the other drones will be able to fly around it -- but if a central UTM system fails then *all* the drones will be left to fly blind without any real way to coordinate their movements and avoid disaster.
However, the bottom line is that there really isn't much drone activity happening at all, on a commercial level.
Predictions of a world where your burger and fries is delivered by drone are a long, long way from fulfillment and even if this does come to pass, our AI technology will have been massively improved by then -- further strengthening the case for autonomy over central control.
There is a huge amount of money at stake in guessing the future in this game. Whoever owns the UTM controls the (unmanned) skies and gets a chance to make a fortune by charging for use of the airspace. "No pay, no play".
For that reason, there's already huge lobbying going on around the world, as key players bump and push to try and become "the" preferred UTM provider. Here in New Zealand we've seen AirMaps convince CAA to support it in limited trials at key locations and elsewhere in the world other companies are cosying up to their own administrators with the same goal in mind.
The really worrying thing about this is that, given the amount of money involved for the winner, I strongly suspect that there will be a lot of "secret handshake, Swiss bank account" type of deals on the table (or should I say "under the table") and that decisions will be made as much on the basis of "what's best for me" as on "what's best for safety".
I've looked at both the AirMaps and the Altitude Angel (two competing UTM prospects) and for the life of me, I can not understand why CAA in NZ h as gone with the former. In my own testing and use I found that Altitude Angel is vastly superior in just about every respect -- yet, for some reason, CAA is going with what appears to be an inferior product.
I'd ask them why this is the case but sadly CAA seem incapable of meeting their own self-imposed deadlines for responding to my queries (a matter I have referred to the Minister of Transport).
Stay tuned, the future looks very exciting... if a little confused.
Please visit the sponsor!
Have your say in the Aardvark Forums.