As drones gained traction, both in the consumer market that has now reached several millions as well as in the business area where drones are really starting to be used and where they are expected to become part of industrial and delivery chains –2.4 million expected to ship in 2023-, the issue of managing their traffic is now becoming crucial.
In the consumer market we have already seen a number of limitations being enforced, as result of problems like presence of drones in airport areas or in other restricted areas. In some countries regulations require drones pilots to take a course and get a sort of “pilot license”. However, nothing has really been made that can compare to the set of rules regulating civilian air-traffic. Yet, what we are facing in the coming years is a traffic density that will go well beyond the traffic density of civilian air traffic (the military one is way lower, generally speaking).
One of the problem is that civilian air traffic control does not scale. Have you ever been stuck in a plane awaiting for the magic slot with the pilot informing passengers that there is a traffic congestion over… and so there will be a delayed departure? Last year there was an average of 106,850 flight a day in the world (44,000 in the North American Airspace), a big number but a number spread over a big “planet” (although unevenly spread).
Managing hundreds of drones flying at the same time in a urban environment requires a different traffic management paradigm, relying much more on autonomous control.
This is being discussed in an interesting post by DroneII.com, a company that in a few years has become the reference point for the drone industry.
As shown in the graphic they identify five components that are needed for the future of drones, a future that is coming closer at an accelerated pace as result of an increased demand for “unmanned” activities, like infrastructure inspection, agriculture, mining, delivery… following the pandemic.
There are agreed standards for supporting drone to drone communications. The new 5G wireless network with its expected dense coverage in urban environment and very low latency would seem an ideal infrastructure and indeed a few Telecom Operators are using drones as a point in case to show the capabilities of their (future) 5G networks. In this area the pace of deployment is not tha critical, since drones are expected to service a (relatively) small area, like a city, and if there is a demand Telcos may deploy 5G. The technology is available, 5G dongles for drones may not be available now but if demand rises they can become available in a few months. Also, the signalling traffic generated by drones is very little in terms of network capacity required so it won’t be a big deal to create the wireless infrastructure (notice that I am only addressing the aspect of signalling required for autonomous traffic management, not the bandwidth that may be required by services, such as drones monitoring the city and generating 4k video streams…). From drones manufacturers point of view, however, there is a need to equip drones with both 4G and 5G since 5G will be limited to a few markets for the next 3-4 years and of course they want to sell their drones to all markets. Hence, we can expect that both 4G and 5G communications will be used to support autonomous traffic management. In case where 5G is not available and 4G needs to be used the applications controlling the traffic patterns will have to downscale the performances, i.e, support lower traffic density,.
There are a number of companies that are developing and offer autonomous traffic management applications (see the most recent map here). There are applications that can orchestrate traffic of hundreds of drones and there are applications that assume independent control. In this latter each single drone is looking for a flight path, and changes it dynamically, “sensing” other drones but taking autonomous decisions. In the former case all drones are reporting to an unmanned traffic control point, one for each urban area, and get direction from it. Of course this requires the registration of each drone onto the traffic management platform.
These applications require regulatory approval, stakeholders support (they should be able to “run” on the drones) and widespread adoption (in the case of unmanned traffic management centre).
Airplanes have a unique identification number and a unique flight plan number. Drones need to have something similar and there is work under development. It is a complex area, it requires the involvement of industry and regulators. If you are interested in the current situation and ongoing work you can take a look here.
With the exception of a few manufacturers that are targeting businesses using drones fleets, most producers have not looked a the aspects of traffic management (all drones are supposed to fly “solo”). However this aspect is now being considered and we can expect the future generation of drones to come equipped with APIs that support swarm management.
Regulators are a most important component and although so far they have been focussing on limitation to drone flight and ensuring the drone pilots are aware of regulation and know how to apply them they will have to start looking into this area. The challenge for regulators is facing a quickly evolving technology that can make regulations established today useless in tomorrow’s environment.
Overall my impression is that we will progressively moving into a framework where drones behave like swarms, applying a very limited, and simple, set of rules ensuring autonomous flight and obstacle avoidance. At the same time the regulation will have to move from controlling the single drone to controlling the swarm. This is an aspect that has been considered by the Symbiotic Autonomous Systems Initiative, now morphed into the Digital Reality Initiative of the IEEE FDC.