EXCLUSIVE - The future of drones – Autonomy, Miniaturisation, Enhanced Endurance
There are few technologies which have captured the popular imagination in recent years as strongly as drones. A drone might bring to mind a friendly flying gizmo delivering pizzas to your doorstep, or a war machine raining death and destruction from the sky.
They are being used for surveillance, rescue missions, surveying and mapping, precision agriculture, insurance inspections and more. New applications are being found all the time.
OpenGov had the opportunity to discuss the future of drones with Mr. Arthur Holland Michel (above) on the sidelines of INTERPOL World in Singapore in June. Mr. Michel is Co-Director at the Centre for the Study of the Drone. It is a non-profit research and education institution founded in 2012 and a part of Bard College, a university in New York state. The Centre conducts research and produces educational content relating to challenges and opportunities that come from the proliferation and development of unmanned systems technology, both in the military and civilian domains. Within the civilian domain, that means both public use, safety agencies, governments, but also private use, whether recreational or commercial.
“The most important thing to note about the Centre for the Study of the Drone is that we are an inquiry driven organisation. We do not have a policy agenda. Our goal is to educate stakeholders about the technology and its implications so that they can then have a public debate and develop strategies and resources. We serve a wide range of stakeholders, in government, media, advocacy groups, the public at large,”Mr. Michel said.
Trends in drone technology
The technology is becoming smarter. Things like obstacle avoidance and detection, autonomous network behaviours, not just having one drone but having multiple drones that can behave in a coordinated fashion, drones that can perform mission planning among themselves, are all derived from advances in autonomy.
Mr. Michel pointed out that autonomy is always on a gradient. It’s not like there is a line between autonomy and lack of autonomy.
“A drone that can do automated obstacle avoidance has a degree of autonomy. In the commercial space, when we talk about autonomy that is commercially available, we are talking about things like automated tracking, automated obstacle avoidance, automated waypoint navigation,” he said.
The next generation of autonomy, advanced autonomy, things like network behaviour, swarm behaviour, things like automatic decisions based on sensor data analysis, so a drone recognises a person’s face and then takes particular action based on that, that is all still in the development stages.
The technology is getting smaller. Mr. Michel pointed out that DJI’s newest drone, the Spark, is small enough to fit inside your pocket. That is thanks to a range of advances in miniaturisation that did not exist a couple of years ago.
There is also miniaturisation of the sensors. So, the platform itself is getting smaller or you are able to do more with the same-sized platform.
“An infrared camera now is small enough to fit on a smartphone. All of that has a direct impact on unmanned systems technology because it means you can carry more capability in a smaller package. Weight is everything when it comes to drones,” Mr. Michel said.
There is an increasing interest in the military in having small drones that can provide squad level support. The US Marine Corps for example recently announced that they will be using 3D printed small drones in theatre in the coming months.
A third trendline is endurance. The technology is improving allowing drones to fly for longer periods of time. With that everyone is sort of waiting for a breakthrough. Lithium ion batteries are fairly heavy and they give a commercial drone about 20-30 minutes of endurance. The industry is trying hard to push that further. Because if the technology has greater endurance, you will be able to do all sorts of new things that currently aren’t possible.
Drones that don’t fly
When Mr. Michel mentioned that there is increasing interest in the military in drones that don’t fly, such as unmanned ground vehicles and maritime drones, boats and submarines, we asked if he looks at all these as ‘drones’ in his research.
He replied yes and explained, “We think it’s useful to consider drones as being not just aircraft. As the technology continues to evolve, the definitions are going to become more blurred. You are going to see drones that can both swim and fly. You are going to see drones that can operate on the ground and can also swim. We also think it’s useful to use this broad definition because a lot of the same issues arise from unmanned technologies across domains. Autonomy, regulations, these are all questions that you see for unmanned ground vehicles or unmanned undersea vehicles, just as much you see for drones in the sky.”
Barriers in the way
There are technological barriers and regulatory barriers and the two are related.
There are some barriers, such as endurance, that are sort of intrinsic to the technology.
If we have a large facility and we want to fly a drone for say perimeter security, it would be impossible to conduct persistent surveillance with a drone can fly for only 20 minutes.
We can get around the endurance problem by having a tethered drone, a drone that is attached to the ground by a thin cable. That provides unlimited endurance but limits the mobility. The drone can’t move around.
Mr. Michel gave another example, “You think about police operations where you are going to have a drone overhead. Unless the operation is less than 20 minutes long, the drone is going to have to come down to ground at some point.”
Finally, there are limitations in terms of autonomy. If someone has invested 60,000 dollars in buying a drone, they would not trust it to execute a mission autonomously, without crashing into something, without messing up. That results in the need for a human operator for the drone.
Then there are regulatory barriers that have to do with the technology. Because the technology is not yet at a point where it is fully reliable and robust enough to do autonomous missions, swarms of drones are not legal in any country.
“So, in most countries, Amazon’s drone delivery programme is not legal at the moment, because for that to happen, they would need swarms of autonomous drones. Until the technology can actually evolve along these trendlines, some of those regulatory barriers will stay in place,” Mr. Michel said.
There are regulatory barriers limiting the size of the drone. In countries like the United States, the drone has to be less than 25 kilograms. If we want to carry a 25 kg package, we can’t do that with a 25 kg drone.
There are some other limitations which exist from a regulatory perspective. Such as you can’t fly at night in many countries. It’s a safety consideration.
There’s a conundrum here. “The regulatory bodies want to see that the technology is safe before they create regulations that enable it to do all of these things. But these technologies won’t be able to demonstrate that they are safe until they are used at a vast scale, which requires the regulators to open up. It’s a catch-22,” Mr. Michel explained.
Most countries are looking seriously at regulations for drones. In countries where there are no regulations yet, there is a really strong imperative for them to build regulations, because they are getting pressure from industry and from users.
International collaboration on regulation
The International Civil Aviation Organization (ICAO), a UN agency recently is looking at unmanned aircraft system traffic management systems. The European Union (EU) has started to develop a proposal for an operation centric, proportionate, risk- and performance-based regulatory framework for all unmanned aircraft.
In the military area, there are serious concerns. Currently military drones always have a human pilot. In the future though, drones maybe autonomous. They may be able to make a lethal decision autonomously without a human operator. So, a drone identifies a target, uses computer visions, makes a decision that it is a threat and executes a kill on that target autonomously.
The UN’s Convention on Certain Conventional Weapons (CCW) has had several annual meetings specifically looking at this issue. They have experts that are actively researching the issue with an eye to potentially developing an international agreement, similar to the international agreements for cluster munitions, land mines, for certain non-lethal weapons like blinding lasers. There has been a lot of international discussion and dialogue (here and here) to look at potentially developing regulations or a convention.
“There is some talk of potentially getting to that point. But there is a lot of disagreement. It’s a very, very complex process to establish a convention like this. It still remains to be seen whether it will actually happen,” Mr. Michel said.