Why Pilots and Planes Will Become Obsolete In The Near Future . . . And What We Can Do About It (Part 1)

March 6th, 2014 by John Petersen

It is not at all farfetched to believe that the pilots and planes that we all know and love are, well, on their way out – that we are at the end of an historical era.  The indicators are all there, both in terms of what we know about the past and by observing the current trends that surround us.

History tells us that everything changes.  So it is inevitable that the present paradigm will give way to something new.  The only question is when and how.  To understand this, we must begin by describing the larger environment in which we find ourselves – providing a context for understanding the other forces that are in play.  As it happens, the context is unprecedented and extraordinary.

We are living in within the highest rate of change in the history of humanity.  Never before has our species (or any species, for that matter) experienced the converging exponential forces that are presently catapulting us toward the horizon. No matter which dimension you choose – technology, social values, agriculture, science, energy, climate, government, et.al. – we are confronted with situations that would largely have been considered implausible as late as five years ago.

Enabled by the global neural system we call the Internet, the increasing interaction within support systems (and our values and perspectives) are rather amazing. Changes in one area ricochet across many others, generating cascading shifts that follow each other with shorter and shorter intervals.   The metabolism of the whole human experience is amplified by the feeding of trends and events on each other, producing larger and larger impacts.

It’s within this context of rapidly accelerating breakthroughs (and the erosion of the legacy systems) that a number of trends have established themselves that will have direct impact on the future of GA.  These weak signals or early indicators are harbingers of what are sure to become larger, converging forces that will usher in a new era in aviation.

Technology

The combination of ubiquitous connectivity, increasing bandwidth, advanced sensors and decreasing cost is assuring that Autonomous Systems (aka drones and unmanned aerial vehicles) will become an increasing larger segment of the global aircraft fleet. The Navy has flown its first drone from an aircraft carrier, the Air Force is having a hard time hiring the number of drone pilots that it needs, and the Marine Corps is already using a drone cargo helicopter in Afghanistan.  Large drone cargo aircraft are already being designed and UAVs are very rapidly proliferating throughout the law enforcement, news gathering and research communities.  The FAA has certified the first commercial drone and forecasts that 10,000 of them will be in the air over the US by 2020.

Cargo Drone envisioned by Dorsal Aircraft Corp.

Cargo Drone envisioned by Dorsal Aircraft Corp.

Artificial Intelligence is on the horizon. Strong AI agents will act like humans – they will research, collect information (from sensors and other sources), interact with other agents and humans and make decisions.  Think of them as a pilot that knows what the weather is, is constantly aware of the state of the airframe, powerplants, communicates with ATC, filed the flight plan, and flies the aircraft. Advanced Voice Recognition will allow the AI to interact directly with humans. Augmented Reality already has the capability to superimpose information from databases located anywhere on the planet onto the synthetic image generated by the AI controller/pilot.  Advanced Materials are also being developed that will have thousands or millions of miniscule computers embedded within them that will signal the state of any aircraft component (temperature, pressure, etc.) on a real time basis to the AI pilot.   It’s not certain when this capability will become commercially available but I’d guess we’ll begin to see applications within a decade. In any case, they are certainly coming.

The integration of these capabilities (and others) present the rather real possibility of getting into an aircraft in the not too distant future, telling the techno cab driver-controller where you want to go and sitting back while it determines the ideal route and then takes you there.  If this seems farfetched, keep in mind that technological advances are more than doubling every 18 months so application in 2020 won’t be just five or ten times better than today but will be over 500 times more capable. Ponder that for a minute.

This kind of explosive development also raises the distinct possibility of the emergence of things like levitation into the civil fleet in the not too distant future.  There are a number of private efforts underway to develop this capability and an application of the technology has been reported in the major aviation press to already be an integral part of the wing design of the B-2 bomber.  In any case, levitation would obviously produce an aircraft that didn’t look like or operate like those we see at our local airport.

(To be continued next month)

John Petersen

John L. Petersen is a futurist, strategist, and pilot. He is a former aircraft carrier based naval aviator, aircraft builder, and author of three books. He founded The Arlington Institute, edits and publishes the free e-newsletter FUTUREdition, and is the chairman of the Lindbergh Foundation.

The opinions expressed by the bloggers do not reflect AOPA’s position on any topic.

  • Shary

    “Strong AI agents will … communicate with ATC, …”
    Now, of course, ATC will be robotized and completely AI about the same time that pilots disappear. And why the need for drone operators (human, one each)? Robots can handle that task just as well. And why worry about the military? Actions will be merely one Terminator against the other Terminator. Humans will eradicate each other by shear overeating and overpopulation …

    • 00298488

      MMMMM, a true “Fatalist” that disallows human instinct as to “self preservation!”

      • Steve W

        Yeah, self-preservation and under bleak environmental conditions that will manifest in brotherly love I suppose. That has happened so often in the past; in less sever global calamities that it is bound to be solved in an equitable manner, again, what are you dreaming. I am sorry, unless population growth is brought under control the hand writing of how it will unfold is on the wall. All this tech is really a distraction, a tempest in a teacup. Maybe drones and pilotless planes will deliver the final solution on an out of control humanity. Dark thoughts, perhaps but not beyond the direction we are heading at all levels. Picture it, the few at the levers of power, decide to do what must be done in their consideration, to be implemented by machines that will not question their conscience in the execution of the process. Like this has never happened before, right, just not on the same level. But I digress; this is a topic about tech not human survival. I just felt compelled to respond to 00298488, but wait, with a handle like that I might be responding to a machine. I wonder if it is in its programming to understand what was being said.

  • aviweiss

    Humans, by their very nature, are frequently inaccurate, imprecise, and inconsistent in motor-skill based tasks, especially ones where “judgement” must be regularly applied in between tasks; Just like elevator operation was fully automated, ultimately all vehicle (terrain, marine, and air) control manipulation responsibility can and should be delegated to systems that will be far more accurate, precise, and consistent than human operators could ever be, even during their moments of greatest proficiency.

    Decision-making and judgement based on dynamic conditions and situations are processing aspects where humans (still) excel over automation, and likely will for a considerable time to come, even with “AI” capabilities that are emerging and on the horizon. While it may “scare” some riders and no doubt anger seemingly-disenfranchised “pilots” who see aircraft manipulation as part of their domain, accepting that “division of labor”, and fully committing to it during vehicle design will enable vehicles to be more efficient, enhance safety, and ultimately have societal “knock-on” effects across a wide spectrum of issues, from safety to efficiency to environmental improvement.

    • 00298488

      After reading all I’m inclined to believe, this infomercial type talk does nothing to enhance younger people to seek a career in Aviation… Not good!! Even the UAV/RPV’s require a human to control them…. We are obsolete, nor will we become same..

      • Lardo140

        The elimination of elevator operators was done for economic reasons. Not because of “frequently inaccurate, imprecise, and inconsistent … motor-skill based tasks, … (or errors in) “judgement”.

        • aviweiss

          I mentioned the automation of elevators to bring in the notion of “push back” that the airlines will receive, much like the elevator-equipped buildings received from the many elevator operators who were disenfranchised by such automation. It was not mentioned to compare manipulating a one-degree-of-freedom vehicle to one with six-degree-of-freedom, which would be a non-sensical comparison.

          That isn’t to say that airlines wouldn’t love to get rid of pilots for economic gain as well. Salaries, workman’s comp, insurance, boarding expense when away from base, reduction in weight (or added payload), reduction in aircraft complexity… all reasons an airline would love to have automated aircraft, in addition to the consistency and precision afforded by automation.

          • Lardo140

            Well you’re right. If the airlines tried (and given the opportunity, they will try) to go pilotless, there will be push-back. But I think there will be even mor push-back from the flying public. I know I won’t get anywhere near an airliner flown solely by computer. (Especially if that airliner’s computer system is running Microsoft Windows. )

          • aviweiss

            There would be traveler push-back, but would be mostly emotional in nature, not fact-based, and certainly not reasonable, given the weakest link at the moment are us human pilots, NOT faulty automation (though poor automation man-machine interface design is another story).

            Much of commercial flight today is ALREADY highly automated, so much so that B&CA magazine had extensive coverage in their February issue related to pilots becoming “children of the magenta”; overly dependant on automation and losing much of their basic airmanship skills, which the Asiana accident so brutally highlighted. With the coming addition of surface radar for taxiing, aircraft will soon readily be able to roll out of gate, taxi to runway, takeoff, climb, fly to destination, shoot a full Cat IIIC autoland approach, land and taxi to gate… all unassisted by human hands.

            Yes, getting ALL the “bugs” and issues worked out from the software and hardware will take additional time, effort, and money, given the stringency of certification requirements and exceptionally low failure rates tolerated (10^-13 I believe but don’t recall off hand), but it most certainly CAN be done, especially given the huge leaps in sensor, processor, display, and actuator design and technology that has been achieved in last 20 years. Once this is done, automation WILL enable the accuracy, precision, and consistency I alluded to above that will greatly reduce the risk of accidents in ALL phases of flight, much more so than is accepted as “small” today with humans still at the helm.

            As a side note, the RTCA DO-178 process required to certify software for “flight-critical systems” is extremely exhaustive, and essentially requires a line-by-line review of code, which is one reason Microsoft has never wanted to be a part of flight-control software systems, even from the early days, since it would reveal far too much of the Windows operating system than they were prepared to allow. We had tried to use the kernel many years ago for an avionics suite I was apart of designing, but they were not interested.

  • Arend Rietkerk

    What Humanity would need first, is robots that clean the bathroom make the bed and other drudgery jobs. Before we need robots to fly planes.
    We could incentivice the Capitalist system a bit more, by raising janitor’s salaries above $100 000 per year.

  • ssira

    This is naive. The variables for manned flight on the private level are enormously more complex than that of driving a car. For the same reason we do not see a car that needs no driver to autonomously navigate the streets from one location to another is a subset of the reasons why autonomous aircraft are not feasible.
    Drones? Sure.
    Automated aircraft? Nuh uh.
    With mother nature, 3 dimensions of travel, passenger medical issues, failed communications, stuck mics, jammed frequencies, negative GPS coverage/reception, errant aircraft, manned aircraft doing stupid stuff, drone ‘pilots’ doing even dumber stuff, deer on the runway, cows on the runway, flocks of birds zipping around airports, and literally hundreds of other variables, I think this article enters a unique realm of ‘naive’..

    • lowapproach

      Agreed. I can’t even count the number of heated debates I’ve had with coworkers over the idea of autonomous google car future where steering wheels are banned. I always tell them the best computer is the one between your ears.

    • Louie Remigio

      Btw I agree with you…even though I almost took a technological position below. In the day to day life of operating an airliner, a lot more goes on than people realize. Much more than maintaining a heading and altitude to a destination.

    • doodahdave

      I’m an active airline pilot and I have been predicting this for years. Aircraft with “fly by wire” (Airbus/Boeing/Embraer) controls are already controlled via computer and would only require the addition of a drone type data link to drastically reduce the number of onboard pilots needed. The cargo carriers like UPS or FEDEX will be the first to install pilot awareness enhancements to monitor flights on long, boring ocean crossings. They’ll sell it to the FAA as a safety enhancement. Then they’ll lobby to have the mandatory relief crew member requirement removed and allow one pilot to sleep/rest if a drone operator is available to augment the crew.

      Once proven, the same technology will migrate to the passenger airlines, and as the traveling public grows in confidence with familiarity of self-driven cars, pilots will become symbolic system monitors.

      • ssira

        @doodahdave. You covered a lot of ground starting with fly-by-wire. Your progression of logic is interesting but flawed. Fly-by-wire was a natural progression from direct cables used by the wright brothers.
        But fly-by-wire or fly-by-cable has nothing to do with drone-type data-links. Drone-type data link can control any aircraft system just like an STEC-55X autopilot can control a Cessna 172 (just a bunch of cables on that).
        So your basis on this point is… well… pointless.

        So, lets talk about data-link control in general. Yes, it is here and can be applied to passenger aircraft. But it needs to be proven fool-proof, immune to data link outages, and impossible to hijack. That public confidence you mention will go quickly to zero and stay there the first time some hacker in his parents’ basement takes a few hundred people on a joy ride over the Arctic (or worse). Data-links are tough for long distance and remote communications – or perhaps you’ve never noticed comm outages on your various devices/equipment/whatever.

        Yes… I too believe the relief crew requirements can/will eventually be discarded due to enhanced technology/communications. These requirements weren’t there years ago and they will be proven unnecessary again.

        But really… big deal… to claim that is a step towards autonomous aircraft flight is absurd.
        And to suddenly discuss self-driven cars because you think relief crews will no longer be necessary… ….? – sorry — but that is a leap that defies gravity,

        • doodahdave

          It’s coming, and I’m not happy about it. Have you seen “Children of the Magenta”?

          My point is that fly by wire systems have already supplanted hand flying skills in providing envelope protection in new airliners.

          As an operator, the challenging part of engineering a remotely piloted vehicle is designing the system to react to immediate changes in flight conditions, fly by wire software does that. Flying the rest of the flight is well within the capabilities of most autopilot systems. Airport approach control and landing could be done by line of sight communications to eliminate satellite lag times.

          Besides, long haul aircraft are only in contact with ground facilities intermittently (position reports-mostly automated nowadays), and satellite communications are already much better than old HF radios. And as far as data link security goes, I’ll bet the military has workable solutions.

          As the public develops familiarity and trust in autonomous vehicles, having only one pilot/operator on an aircraft will not seem so frightening.

          I’ll bet no one remembers when the Flight Engineer position was eliminated. Now two pilot aircraft are the norm, are single pilot airliners inconceivable? Even with a looming pilot shortage?

          We’ll see.

          Especially if the ticket is $5 cheaper on Priceline.com.

          • mf

            You are probably right that fewer pilots may be allowed to fly a large aircraft with enough automation. However, it is not clear to me what happens when the automation fails. I think that complete elimination of a pilot Is a long way off. Perhaps a more pressing problem is how to keep humans alert and properly trained. I still can not get over the Air France accident, or Buffallo for that matter.

  • Louie Remigio

    My questions to the author is….why do you limit this to pilots and airplanes? Why not doctors and medicine(google ‘robot doctors’) ? Accountants? Architects? Software engineers? Futurist and strategic article writers? The list goes on and on. Artificial Intelligence coupled with big data coupled with exponential increases will affect every profession. Isn’t a computer able to write sports articles now? It’s not unfathomable to have computers program themselves, design buildings, automate accounting for companies. These all bring up larger philosophical questions as well. I once heard a commentator say that machines and technology will rise and become the next stewards of the universe.

    In the meantime I’ll go outside and asides from seeing people using the awesome power of technology in their cellphone to look at Ellen Degeneres’ selfie’, life still looks like 1975!

  • Daniel Lee

    Let’s say you had a valuable diamond and you didn’t want it stolen. If you try to protect it with an alarm that would be easy to defeat. But have a human guard watching and it is almost impossible to steal, because the human very quickly adapts and learns, where a machine that “learns” must make random choices to find out what works and what doesn’t and that’s too slow of a process.

  • Daniel Ford

    Why do folks speculate about pilot-less aircraft but overlook the numerous safety and economic benefits of using drone technology to assist human pilots?

    For example, in the event of a suspected hijacking or severe illness, a pilot or crew member could electronically transfer control of the aircraft to a remote pilot. Or on a long flight or a flight to an unfamiliar airport, a remote pilot could act as a monitor, translator, or a third set of eyes in the cockpit.

    The economic and safety case in favor of electronically augmented crews is far stronger than the case for pilot-less aircraft in that it need not have the numerous and costly levels of redundancy that human transportation requires, existing “off the shelf” technology could be used, and it would be retrofitable to the existing fleets.

    • karlschneider

      How would transferring control to a remote operator reduce the risk that a hijacker could start killing passengers if his demands are not met?

      • Daniel Ford

        A hijacking is an exercise in control. Anything that potentially interferes with the absolute control a hijacking requires would be a deterrent, much in the same way a car thief would opt to steal a car without a remote “kill” switch than a car with one. At the same time, the same technology can be used to mitigate other risks that are statistically a greater threat to safety than hijackings.

        • karlschneider

          Your analogy is nonsense. A person bent on hijacking an airplane full of public citizens is nothing like an automobile on a city street. You failed to address my question….abysmally.

          • Daniel Ford

            If you do an online search, I don’t think you’ll find a single instance of an armed airline hijacking where the bad guys “allowed” the aircraft to continue to its scheduled destination.

            Ergo, having the ability to transfer control of the aircraft to an external location at the first sign of trouble would disrupt a common component in any hijacking plan and would act as a deterrent.

            At the same time, the technology could assist the crews in their normal duties and in other emergency situations making it very cost effective.

            See: http://en.wikipedia.org/wiki/Airline_hijacking

  • karlschneider

    Wait for the digital PA announcement: “Welcome aboard on your pilotless flight, everything is controlled by computer. Nothing can go wrong
    go wrong
    go wrong
    go wrong
    go wrong…

    • Edouard Kohler

      I do not want to be part of this world. Computer should just be a tool to help not dictate and control our destiny.

  • Andrew333

    Why? Pilots are cheap.Where is the financial benefit?
    Adding layers of unnecessary and fallible technology and complexity to something does not make sense.
    In the end,
    1. You cant take a human out of the loop without adding considerable risk. AI? Not with anything or anyone I value on board. Computer geeks have no idea how incredibly complex and interactive the aviation environment is to model. They completely underestimate the difficulty of the task. Who is going to pay? One crash and their business model collapses due to a lack of public confidence.
    2. If there is a human in the loop, then they need to get paid. Thus all financial benefits of remote piloting go out the window. Doesn’t matter if they are in the cockpit or in a bunker in Arizona. The cost of piloting will need to be paid.
    3. If the human-in-the-loop is not in the plane (but in a bunker in Arizona), then they lack situational awareness and the cost of the computer links etc will exceed the cost of putting the pilot in the cockpit.
    4. The larger the aircraft, the cheaper pilots get. Pilots are cheap anyway. The more passengers that pay for each pilot, the cheaper they get. When it comes to flying passengers around, it will always be cheaper, safer and simpler to just pay a pilot!
    5. The case for AI or drone/remote piloting only makes sense with small air vehicles with no people on board, where the pilot is a large % of the AUW of the aircraft. Once you put a person on board, you need to provide space, life support systems, seats etc and the cost skyrockets. Keep the humans out of the drone and you can fill it with fuel and sensors and make the drone much more cost effective mission platform.
    Conclusion. Thus, ‘small to medium, remotely piloted drones with no humans at all on board’ is the only concept that stands a chance of ever having a realistic business model. Every other AI or drone like concept will have a fatally flawed business case.

  • http://laird.zchs.org/ G. Fisher

    Uh-huh. Right after the widespread adoption of autonomous rail service.

  • sfrisius

    Of all the terms used for drones, UAVs, etc., the most accurate one is RPA. Remotely piloted aircraft. There’s still a person operating the thing. As a 20+ year Air Force pilot, I’ll share a little secret. My friends that are flying RPAs are getting hired at starting salaries significantly higher than I can hope to see as an airline pilot for many years, even at a major carrier. And reliable, secure data links are extremely expensive. These costs will eventually decrease, but we have a long way to go before an RPAirliner is financially viable.

  • Tomas Adorno Montes

    Why would private aircraft be manless?We already have drones and future aircrfaft could be automated without pilots.Air taxies but would it be exonomically feasable to have private planes pilotless ?Who is going to pay the bill for a computer to fly you own plane?Where is the fun on flying at?

  • Tomas Adorno Montes

    Bill gates said that computers will make a paper less society?jajajajajaj it turns out that companies and banks have more paper work now .

  • Tomas Adorno Montes

    I believe that composite materials will make it more feasable for private owners to fabricate their own aircraft.The experts in computers like Bill Gates said that computers will make paper documents alsolete ?buhu now we have more paperwork.

  • Tomas Adorno Montes

    i believe that Composite materials will give the pilots the opportunity to create their owns aircraft too

  • Daniel Winkelman

    All very interesting, intellect stimulating, fascinating to mull over. Not a shift that will happen that quickly. Yes, technological hardware capability may progress to 500 times its current capability by 2020. That does not mean *reliability* will progress as far or as fast. Nor will coding (which is equally, if not more important to a well-functioning system). Almost everyone in the industry is familiar with the fact that hardened electronics have performance that is 5-10 years out of date. Why? It takes that long to build-in and test reliability. It’s common complaint amongst the programmers that they must deal with memory and processor limitations in avionics. First and foremost, avionics must be bulletproof. The hardware must handle electromagnetic interference (EMI), conducted emissions (interference on the wiring), wide temperature swings, and intense shock and vibe requirements. Every logic path must be deterministic. Processor loads must be kept within strict limits. Memory use must be far more tightly regulated than in a personal computer. Coding standards are far higher. Case in point: Windows is not allowed in any certificated aviation use. The code is simply not robust and reliable enough.

    If your personal computer crashes, you may have lost a few hours worth of work. If the primary control system of a UAS crashes, you have much bigger problems… hence all the reliability and durability requirements.

    The major point is, all this takes time. Add in the fact the FAA takes a LOT of time. They will only have an integration *roadmap* in place by 2015, if they can even hit that target (UAVs themselves will not necessarily be allowed in the NAS starting right then). Integration itself will be slow and staged, with lots of stopping and checking for safety considerations. Humans will be in the loop for many years. After that, humans will be “on the loop” for many more years. I’d put money down that it’ll be at least 30 years before the first passenger-carrying pilotless airline flight. My gut feel is more like 40-50 years. I can guarantee you the biggest challenge won’t be advancing computing power… it will be advancing safety, reliability, redundancy, ruggedization, and verification & validation processes. Yes it will happen. It will just take a long time to do it safely.

  • Pingback: click here