Beyond military drones – the future of unmanned flight


June 29, 2013

The modified Jetstream that flew 500 miles (800 km) under ground supervision (Photo: BAE Systems)

The modified Jetstream that flew 500 miles (800 km) under ground supervision (Photo: BAE Systems)

Image Gallery (4 images)

In April of this year, a BAE Systems Jetstream research aircraft flew from Preston in Lancashire, England, to Inverness, Scotland and back. This 500-mile (805 km) journey wouldn't be worth noting if it weren't for the small detail that its pilot was not on board, but sitting on the ground in Warton, Lancashire and that the plane did most of the flying itself. Even this alteration of a standard commercial prop plane into an Unmanned Aerial Vehicle (UAV) seems a back page item until you realize that this may herald the biggest revolution in civil aviation since Wilbur Wright won the coin toss at Kitty Hawk in 1903.

The Jetstream flight was conducted as part of the Autonomous Systems Technology Related Airborne Evaluation & Assessment (ASTRAEA) program, which is a UK industry-led consortium aimed at developing unmanned aircraft that can operate routinely in civilian airspace. It’s one of hundreds of UAV projects around the world, but what’s notable about it is how the use of a passenger plane blurs the line between quadcopter-with-a-camera jobs and full-blown airliners.

The UAV revolution

For many people, UAVs came out nowhere. In the popular imagination, they started out as exotic reconnaissance aircraft in the early days of the Afghanistan War that have grown in numbers and sophistication until they’ve turned into experimental combat aircraft taking off from aircraft carriers. In fact, UAVs have come from a number of areas aside from the military. Hobbyists have made their contributions, scientists as well, and, of course, aircraft engineers.

In many ways, UAVs are a bit like computers. First they were rare and then they were everywhere. Whether they’re tiny quadcopters that fly within a few hundred feet of their operator or huge winged affairs piloted via satellite thousands of miles from the joystick, UAVs are taking to the skies in ever-increasing numbers for a a variety of applications.

One obvious area for UAVs to move into is police work. Take away the missiles and hunting for bad guys on a civilian street is very similar to hunting them on the battlefield. The challenges are virtually the same and transferring the technology from one sphere to the other is relatively straightforward. An eye in the sky could be used for general law enforcement, border control, sea lane monitoring, traffic control, crime scene photography, searching for missing persons, and combating drug trafficking.

The FBI has admitted in recent congressional testimony to using UAVs in the US on a number of occasions and some US police departments have been granted permission by the FAA to operate them. In Canada, a UAV is credited with saving a man’s life. However, other locations, such as Seattle, have abandoned plans for their own drones due to the fierce debate over the use of UAVs in law enforcement.

The question isn't just whether drones should be used at all, but also how they should be used. Should they only be used in extraordinary circumstances, such as hostage situations? Should they be used routinely like a CCTV camera on a pole a hundred feet tall? What sort of oversight is needed? How can privacy issues be addressed? When does policing go over the line to surveillance of the public?

Beyond law enforcement, civilian UAVs in the US are almost non-existent. The FAA has only one rule concerning UAVs, which is that they can’t be used. The agency will authorize UAVs on a case by case basis, but has issued less than three hundred permits. Most of these are for law enforcement with the remainder going to university research programs and pilot projects, such as monitoring wildfires. Even then, most UAVs must operate under severe restrictions, such as remaining in sight of the operator at all times and at very low altitude.

Almost limitless applications

Similar restrictions apply in other nations to different degrees, but UAVs with various degrees of autonomy are already being used and the potential for the new technology is already evident. One area is in agriculture, with Japan having used UAVs for seeding and crop dusting for over 20 years. With their small size and ability to fly or hover very close to crops, UAVs can also be used for what is called “precision agriculture.” UAVs, especially when equipped with infrared imagers and other sensors, can be used to monitor livestocks and crops, detect diseases, determine plant ripeness and schedule harvesting. They can also spread poisoned bait for vermin, such as fire ants, mice and rats with minimal environmental impact.

Other areas where UAVs might show applications is in meteorology with swarms of small robot planes taking up the dangerous task of storm chasing as well as being able to take measurements in situations where using conventionally piloted aircraft isn't feasible.

Another job is prospecting with drone aircraft conducting surveys for oil, gas and minerals. It goes without saying that UAVs are excellent platforms for cartography and geophysical and photometric surveying, allowing an aerial perspective that archaeologists, for one, could only dream of a few years ago. They can also be used for data relays or for inspection and maintenance of bridges and other structures.

But the real question is, are UAVs any good at delivering pizzas? At least one company has tried this as part of a promotion, but a restaurant in London did them one better with flying waiters to deliver sushi burgers. These may be novelties, but they illustrate the almost limitless potential of the technology.

The US Army is already using unmanned helicopters for hauling cargo in Afghanistan and it may not be that far off before we see them taking up some of the load at home from trucks. In the meantime, there’s always the UAV as personal assistant and dance partner, which would be doubly impressive if it was mind-controlled.

All of this seems like a technology shooting off in a thousand different directions with every sort of fixed wing or copter imaginable whizzing about in a chaos of invention, but the UAV revolution is much simpler, though with great implications. It’s less a matter of what’s in the UAV than what isn't.

Removing the pilot

Technology often advances in lurches. Computers, for example, were figured out theoretically as far back as the 1830s, but it took the invention of the microchip to send them off and running. Many devices that we take for granted were once clumsy and expensive until plastics made them cheap and easy to manufacture. The same is true of the emerging UAVs. It isn't simply a matter of taking the pilot out of the machine, but also exploiting the tremendous advantages that removal implies.

We tend to think of a pilot as being indispensable for a plane, but in many respects, he’s a liability. A human being takes up space, needs somewhere to sit, air to breathe, the right temperature and pressure, and needs readouts and controls to operate the aircraft. All of these take up space, add weight, and use power.

Worse, humans are rather fragile. Acceleration at five times the force of gravity or “Gs” can render an unprotected pilot unconscious. Even two Gs can cause a blackout if it’s in such a direction that it causes the blood to rush to the head. You can do a lot to prevent this with proper seating and special inflatable trousers to keep the blood from pooling in the feet while making a fast bank, but there are limits and those limits are reflected in the performance of the aircraft.

Now imagine taking the pilot out. There’s no more need for life support, seats, control systems, readouts or even something as simple as a latch that opens from the inside. All that weight and space is saved. This allows the UAV to be much smaller and lighter. A perfect example of this is the hobbyist’s quadcopter. Being remote controlled, it doesn't need to be able to lift the pilot, so it’s possible to make a drone so tiny that it rivals a hummingbird, remain in the air as long as the fuel holds out, and go where it would be impossible or far too dangerous to send a person.

And it’s not just size. For now, engineers are happy if a combat drone like the X-47B can operate as well as a manned aircraft. But what about when they realize that there’s no pilot inside that needs protecting? The UAV of the future will be able to accelerate and turn in a way that would be lethal for a human pilot – and may not even be possible for one on the ground to control without an onboard computer’s aid. It may be that the Top Gun of the 21st century is an artificial intelligence.

From autopilot to UAV

All of this brings us back to the Jetstream test. With all these potential applications and advancements, what does a flight from Lancashire to Scotland and back have that others don’t. Quite simply, it illustrates that the civilian UAV revolution isn't on its way, it’s already here. That’s because what ASTRAEA did with the Jetstream wasn't about fitting it with something radical. Instead, the engineers took the technology that was already in the Jetstream and took it a step further.

When someone mentions “UAV,” it brings to mind images of quadcopters with batteries that last less than an hour or motorcycle-sized helicopters, or maybe military drones like a Predator, which is about the size of conventional plane, but looks like a blind monster. In each case, it seems as if the UAV is playing catch up with where the aerospace industry was decades ago.

In fact, what has really happened is that conventional aircraft have been developing in such a way that the difference between a civilian plane and a UAV is almost a matter of perspective.

Look at the ASTRAEA Jetstream. It is a Jetstream small twin-turboprop airliner. It isn't an airframe that’s been gutted and all the working bits replaced. It is a production Jetstream with some new technology added to it. Some of these included autonomous systems and a very robust communications system that allows a pilot on the ground to not only control the aircraft, but to also keep in touch with air traffic control, yet the key in the Jetstream’s case was its detect and avoid capability, which allowed the Jetstream to detect other aircraft, keep a proper distance and obey flight regulations while doing so. This last bit is vital because UAVs make other pilots nervous.

But what’s the difference between a Jetstream turned into a UAV and a regular one? The answer is: Not a lot. And the same is true of many newer aircraft. If you've flown in a modern, long-haul passenger aircraft you've already ridden in what is almost a UAV.

This isn't a recent development. It goes back to the earliest days of aviation when designers were already trying to automate aircraft. Early planes were a nightmare to fly and required constant, unerring attention on the part of pilots or reacquaintance with the ground would happen very fast.

By the 1930s, pilots had an alternative to wrestling with a joystick every second while staring at the horizon. The first autopilots had been developed using simple feedback mechanisms and gyrocompasses that made the plane fly straight and level and even maintain a course heading for hours on end without human intervention. Today, these systems have become computerized and incredibly sophisticated. While smaller aircraft may be restricted to simple autopilots to control roll or keep them flying level, larger aircraft have autopilot systems that can control takeoff, ascent, cruising, descent, approach, and landing. In the not too distant future, they will even be able to handle rollouts and taxiing.

This isn't surprising because autopilots aren't just a form of mechanical steering. They can control a plane’s throttles, determine its position using GPS, dead reckoning and radio beacons, and balance the plane by automatically pumping fuel from one tank to another. Furthermore, many modern airliners use fly-by-wire systems instead of direct hydraulic linkages. In this set up, the plane’s computers aren't just an adjunct to the pilot, they are an integral part of the controls and, in a sense, the pilot isn’t so much flying the plane as telling the computer what to do. Add in the situational awareness and ability to be directed from the ground that the experimental Jetstream has and you've got a UAV.

The future

So, are we going to be getting on airliners soon that have as big a crew as an elevator’s? That is, none? Probably not. The problem with civilian UAVs is that the robot airplane doesn't exist in a vacuum. It has to deal with other planes and the air traffic system that directs them. The air traffic control system used around the world today was built around pilots, not robots, and its technology and procedures reflect this. A controller can’t talk to a robot. A robot can’t talk back. It can’t respond to spoken instructions. That may seem obvious, but it reflects the fact that air traffic relies very heavily on human judgement and interaction rather than mechanical systems. Furthermore, humans and machines work differently and, for example, landing maneuvers that are best for pilots may not be the best for UAVs.

Another reason why UAVs would have a hard time is that there are a lot of aircraft that don’t use any sort of autopilot. Older and smaller planes may use nothing more sophisticated than the seat of the pilot’s pants and even small airliners of fewer than 20 passengers on short duration flights with two pilots may not have one. If every plane was equipped with GPS, as is planned for the future, and ADS-B tracking technology, then air traffic procedures could be designed so that UAVs and manned aircraft can mix safely. But so long as human piloting dominates and the sort of time delays and need for human judgement is required, UAVs flying over a couple of hundred feet or weighing more than fifty pounds haven’t much of a look in.

Does this mean that civilian UAVs are a regulatory and technological dead end? Hardly. One effect of the increasingly automated plane is that air cabin crews have shrunk remarkably over the past fifty years. At one time, any large airliner would have a pilot, co-pilot, navigator, engineer and radioman. Today, there’s only the pilot and co-pilot and there’s even talk of eliminating the co-pilot. However, pilots don’t make up much of the cost of running an airliner, unions wouldn’t be happy with eliminating them, and the public don’t seem very keen about flying alone with no other company other than an automated drinks cart. So even a plane that can fly itself or under ground supervision may have a human pilot aboard for emergencies and reassurance.

What this means is that civilian UAVs of the airliner variety may be similar to the military’s man-optional aircraft. One trend in military aircraft design is to make combat and transport aircraft that offer the option of accommodating a human pilot, but isn't necessary. The idea is that, for example, a fighter plane might have a pilot onboard when going on a mission, but fly itself when being ferried to a new base. A similar system could be used with civilian aircraft that might have pilots aboard with passengers, but fly themselves when empty or used to carry cargo over water and landing at coastal airports.

Air taxis

Another way in which we might see the UAV move gradually into civilian life is at small airports. Currently, air traffic is a nightmare of centralized air traffic control systems moving fleets of aircraft between hub airports that are stretched to capacity. Aircraft equipped with UAV technology could change this because such craft have the potential to talk directly to one another and operate with a high degree of autonomy instead of waiting for air traffic control instructions.

What this would mean is that as the technology becomes more widespread, air traffic control becomes much simpler and small local airports would be able to do many of the things that only a major one can do today. Imagine a city the size of London, which needs half a dozen major airports to operate, being served by one large airport and a hundred small ones acting like air taxi stands and you get the idea. Indeed, it may be that the first pilotless passenger plane may not be a huge airliner, but a little air taxi that ferries around only a handful of people.

Of course, this sort of a future depends on a lot of technological hurdles being overcome. As was seen in the Jetstream experiment, UAVs have to see and avoid other aircraft. They need to operate with piloted craft, and they need to be supported by an air traffic system that plays to their strengths instead of against their weaknesses. Worse, UAVs need massive amounts of bandwidth, which the military has, but civilian air systems lack.

Finally, there’s the fact that fully autonomous devices aren't completely trustworthy. Even an autonomous machine needs a human for when things go pear shaped. In an emergency, automated systems can do a lot, but sometimes human improvization on the spot is key.

Ultimately, what the world where UAVs are part of everyday life looks like will depend on the decisions we make. Will we have flying robot baristas delivering lattes to our tables? Will we be stepping aboard a crewless airliner as comfortably as we do an automated train? Will we have our own personal assistant following us around like Tinkerbell updating our Twitter accounts throughout the day? Or will we be ever on the lookout for spying drones? Only time will tell, but the technology is moving at such a rate that these are questions we need to start asking ourselves now.

The video below outlines the ASTRAEA program.


About the Author
David Szondy David Szondy is a freelance writer based in Monroe, Washington. An award-winning playwright, he has contributed to Charged and iQ magazine and is the author of the website Tales of Future Past. All articles by David Szondy

Good article, but there's one specific area where pilots are needed. That's when the s*** hits the fan. In November 2010, a Qantas A380 had one of its engines explode, which blew a hole in the wing and caused other serious structural damage. The aircraft was saved by its pilots (that's according to the Australian Transportation Safety Board, who are not given to hyperbole). Until computers can adapt to unforeseen circumstances, it will be necessary to have humans in that cockpit. After that, who knows?


It is simple to stop UAVs. Outlaw ICE Engines, GPS, Computers, propellers, electricity, jet engines and common sense. Voila! Problem solved.

Robert Walther

The next step in passenger is rather obvious. A secure fallback control (sort of a dead man's switch) in case an airliner's crew becomes incapacitated. OTOH the possibilities for abuse of this technology practically write themself, making layers upon layers of security measures even more necessary.


Normally I'm an avid supporter of technology and innovation. The main problem with this concept is the human element. If unmanned aircraft usage becomes commonplace I think (based on decades of observation) that corporate and government bean counters will start trimming costs and insisting on a level of profitability that will open windows for security breaches in command and control systems. The idea of ground-based telemetry & uplink infrastructure scattered widely also invites intrusion attempts since they will likely be as well protected as current utility power sub-stations and telecom infrastructure, which is to say not at all. Then imagine large aircraft as targets of a hacking takeover. Large unmanned refueling aircraft? Ummm....Sounds like both a disaster-in-the-making and at the same time, a great business opportunity for the makers of Close-in Weapons Systems when major strategic, utility and population centers start requiring land-based CIWS installations - just in case...


On the other hand, if no humans are on board, the plane and its payload can just be written off. All you need to care about is to direct the impact away from populated areas. Those guys are very pragmatic about things, it's a huge market...

On a side note, when flying with a certain French airline, I'd rather have a machine controlling the plane than one of their pilots. But that's just my personal view. ;-)


And you thought a few fanatics with box cutters could do some damage. Wait until it only takes a few key strokes to exponentially take it to the next level! Just saying...

Norman Bouchal

I see cargo deliveries as the most likely first application.

When I see video of convoys in Afghanistan creeping forward at 5 mph and stopping frequently to defuse the IEDs, I wonder if an automated delivery with a smart guided parachute drop to an IR ground beacon might not be a safer, and possibly cheaper alternative.


In 89 a DC-10 lost all hydraulic pressure when its number 2 engine came apart in flight. The assumption at the time was that under those circumstances the crash would kill everyone on board. The pilots doing something that had never been seriously suggest before managed to get the plane low slow and traveling almost parallel with the ground before crashing on the runway where they were trying to land. To everyone's amazement there were a large number of survivors.

A Colorado ANG A-7 suffer a major control failure and the pilot was advised to eject. The pilot a man with a wife and child(ren) saw the houses under him and chose to stay with the plane and try to find some uninhabited real estate to crash the plane into. After discovering what his limited control responses would do managed to fly back to Buckley AFB and successfully landed catching the emergency arresting wire.

I don't think drones could do something similar to either.


All this "Hero avoids crashing into a school/ houses etc" is a red herring. ANYONE in charge of a defunct plane would ALWAYS aim for a soft landing if one is available. It is common sense to avoid crashing into a rather solid brick building, no matter how many occupants are in or how many wives and children the pilot has. Think about it. ;-)


And then we have the multitude of crashes that has come from human error and/or panics.


Drones yes, commercial no.

If you paying someone to fly the plane put their butts in the plane. How does it make sense to have someone on the ground flying the plane and not in the air?


Safer? So there are actually people out there that actually think this makes flight safer?, If you actually think about it presents more problems then it would answers.... unless you represent an "organization" that prides itself on having complete control over everything and would deny having any involvement when anything went wrong.

Its ridiculous to say this is as big as an impact as the method of flight being developed to begin with. Its no different then adding another chair to the cock pit or even an auto pilot. They eliminate the pilot from making their own choice, who could or would ever consider that to be a good thing. If you're afraid of the pilots and their ability then maybe we should put PUI (piloting under the influence) locks on the cabin/cockpit doors?

Here's another thought: Ever wonder what would happen if the controls or electrical instruments were damaged? A human could land the plane, A program/robot is going to send someone an email telling how many people died. They can't even keep them in the air with the current technology and they want to put more on the thing?!

I will never fly on plane piloted by a robot, ever.

Maybe they hope by not having to pay pilots they're be able to save themselves from going bankrupt again....

Any way you look at it, this has bad written all over it.


re; Podzak

There is no reason to believe that the failure on a drone would not affect its communication system.

The Boeing 737 a very safe plane by any reasonable standard. But a small number (Best guess, under 50 possibly less than 10 out of over 7000 airframes) of the rudder controls had a problem that has been fixed. The hydraulic valve controlling it was a complicated double valve designed to keep the rudder functioning even in the event of something clogging one of the pathways but if the two main parts of the valve were at the extreme limits of design tolerances in exactly the wrong way and the valve was cold soaked (it is very cold at efficient flight altitudes) And then hydraulic fluid that had gotten very hot by spinning in the pump not being allowed to go anywhere was sent through the rudder control valve the rudder the fluid could find a pathway to put the rudder hard over in the opposite direction of the pilots input. Pushing harder on the pedal did not help. There were two crashes but several other events of the control reversal taking place. If the pilot pushed the opposite pedal normal function would be restored. The exact problem is unlikely to occur again but something similar probably will and I think the odds of a human pilot finding the solution is far better than a computer finding it.


@ "...even a plane that can fly itself or under ground supervision may have a human pilot aboard for emergencies and reassurance." For emergencies? The human is invariably the cause. Just ask Sully Sullenberger. And a computer can say "Brace for impact!"

Dave Brough

Having read this article has made me realise how boring flying as a 'commercial pilot' will be in the future. Sitting in a chair totally removed from where the action is will be the ultimate in 'job sanitation', no more hands on flying, no more panoramic views to behold, no more seat of the pants flying sensation. I learned to fly when you felt the wind in your face and you had to know how to navigate using charts and your brain, you had to feel the aircraft and the air it was in. You as pilot in command, were responsible for not only your life but also those around you. Being a computer geek at a console takes that aspect away from you, remember the scene from "Flying High II' where the kid crashes the shuttle, swears and just walks away? Real pilots on real flight decks don't have that option if the poo hits the fan. Hell I even find my helicopter flight simulator boring compared to flying the real thing.


Given how current civil aviation is 'hard wired' to having pilots aboard, my guess is that pilotless operation will start with cargo operations to and from airfields not used by passenger craft. Maersk might be very interested in operating pilotless cargo carriers across the pacific for example and departing from a relatively remote Chinese airfield and landing at a remote field in the US Southwest would be OK. They could use newly dedicated routes and reserved altitudes and conflict very little with regular traffic. Same would work from South Asia to Remote fields in Spain, for example and from there to either North or South America. After proof of concept, specialised aircraft or versions would quickly evolve to leverage savings out of the absence of humans on board. At some point Intra-continental service would certainly happen.


The problem is, KevinR, that those of us that fly for fun are vastly outnumbered by commercial interests. Probably end up that we will not be permitted to fly when the UAV's have taken over...............

Martin Hone

Just on the concerns that a computer can't act in an emergency...

With the way Many modern aircraft are flown using fly-by-wire, if the computer(s) were to crash (triple redundancy) the pilot would have little or no control of the plane as it is.

Aircraft control architecture is moving more and more to reconfigurable control allocation systems, where the plane knows what manoeuvre is required (from the flight planning and routing software, with all alternates programmed in just as with a regular flight plan), the aircraft can know what actuators and sensors are available, and regularly test them for functionality, in the event that a sensor or actuator becomes unavailable, the control system immediately reconfigures for NOT having that system available, and optimises for the flight condition.... The actual action taken by the aircraft may in fact be quite a bit more sophisticated than expected, and more precise than a human pilot. Just think that the computer doesn't think in terms of engines, flaps, slats, elevators rudders, spoilers, brakes, reversers and ailerons, but independent control surfaces which provide lift, drag, thrust and torque which can modify the aircraft trajectory...

Imagine the computer being able to extend the kruger and fowler flaps on one wing, to act as (low speed) roll devices while the elevators convert into ailevons for additional pitch and roll control, and the air brakes/spoilers can independently give lift, drag, yaw and roll control. The computer is (would be) able to use the controls in ways not intended and produce a response that the pilot would not be able to with the (4-axis) traditional control inputs. (In the event of total aileron failure). The computer can control control in excess of 18 actuators on a typical twin engine transport. Both pilots using hands and feet couldn't provide that sort of control input, not even if the flight engineer was still on deck.

While this may be a bit beyond the health assessment and control methods of most current aircraft, the military know how it works as they have had fly by wire for many years (even landing a fighter jet with most of a wing shot off, without the pilot realising, just by reassigning the stall speed from something like 130kt to 190 kt and using ailevators and rudders for total control).. In the military, once all systems fail, the pilot punches out and the hardware is written off, with civil operations the technology is becoming more mature, and in a worst case, just pop the steerable recovery chute and drift to a pre-determined alternate landing place (using a computer system unrelated to the normal aircraft control loops, therefore immune to any problems which may be plaguing the aircraft).

It may take another 10 years to get past roadblocks in the bureaucracy. The future of air transport will be semi autonomous (High level supervisory pilot in the loop). Just as space exploration was in the 1970's.

PS, I love flying it rocks... But a flying sardine can is less interesting than a down town bus.

Post a Comment

Login with your Gizmag account:

Related Articles
Looking for something? Search our articles