When Will a Drone Take Your Piloting Job?

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

I've been doing some instrument instruction with a young pilot who's set on working his way through the ratings to an eventual professional flying career, probably in the airlines. I'm occasionally asked what's the best way to do this, but if I'm honest, I have to admit the real question I weigh is whether to do it. That doesn't necessarily relate to the nature of the job itself.

Not that long ago, maybe 20 years, the usual path was to earn the private, instrument and commercial, build some time by instructing and then move on to a charter or freight job. Or do a stint in the military. From there, the choices bifurcate again, offering the world of corporate flying or the scheduled carriers. It still works that way, but there's less flight training going on and thus fewer opportunities in that field. In the military, the demand is for drone pilots, not butts in cockpits.

We keep reading about an impending pilot shortage and while I think there's truth to that, what I'm not sure of is how deep the demand will be and if it will be sufficient to put upward pressure on salaries, at least at the regional level where I couldn't, in good conscience, advise anyone to seek employment.

But there's a longer term trend I think anyone contemplating a flying career has to consider: the inevitable impact of automation and autonomous flight. A young person starting a flying career today at, say, 24 years old, will have a 40-year haul in the workforce. Four decades takes us to 2054, by which time the state of automated flight won't look anything like it does now. There's a veritable tsunami of demand coupled with rapidly advancing technology that promises a rapid evolution if not a revolution in autonomous or remote-controlled flight. Last year, in the U.K., BAE conducted its first conceptually unmanned flight of a Jetstream 31 from Britain to Scotland. It's experimental, of course, but the project continues.

Think where it's likely to be in 10 or 20 years and ask yourself if you think autonomous passenger flights will never happen. I think those who are skeptical tend to judge the question emotionally through the prism of current passengers born in an age where pilots were vested with almost God-like skills and authority. The next generation of ticket buyers might not feel that way. In fact, even as one of those dinosaurs brought up to believe aviation meant pilots in the cockpit, I'm not certain where I stand. I can't say I wouldn't get on a pilotless airliner tomorrow, if one taxied up to the jetway.

So, not to be a buzz kill about the thrill and romance of being a steely-eyed aviator, but anyone looking into the distance should think about this. However much you'll invest in obtaining the training and ratings--$50,000 or $100,000, or whatever—what's the shelf life of the skill likely to be? Pilots won't get booted from the cockpit overnight or even over a decade, probably. But just as the military has replaced piloted flight with remote flight faster than many anticipated, so might commercial flight go the same way, spirited along by advances in processor power and remote datalinking, further fueled by economics that will make that technology cheaper. First, it will be with freight, but inevitably, passengers will be flown in remotely piloted or autonomous aircraft. 

After I wrote this initially, I found this interesting site that features open-ended bets on the future of technology. The bet--which is a pro and con about pilotless airliners--has then-Google CEO Eric Schmidt putting a longer-than-50-years timeline on pilotless commercial flight, while Microsoft's Craig Mundie sees it happening by 2030. That's just 16 years away and a bit too soon, I think. Part of Schmidt's argument for more than 50 years--after 2064--is that it will take the FAA that long to certify such technology. That's an interesting and insightful comment with some merit coming from a commerical pilot who flies his own Gulfstream V. But it also ignores this fact: In the face of overwhelming economic and political pressure, the FAA caves. It happened with GPS and you're seeing it happen with UAS regulation. The likes of Boeing and Lockheed Martin bring big guns to the political table, as do their customers. The FAA will resist, to a point. 

I wouldn't say any of this argues against a professional pilot career. There are still interesting times ahead for pilots sitting in cockpits. But if I were in it for the long haul, I'd have a backup in mind—either allied skills related to autonomous flight or a degree in something that isn't aviation. This is nothing new. I know airline pilots who are engineers, accountants, business administrators, IT experts or marketers. The standing theory was that this provided a safety net against the unavoidable ups and downs of the airline biz. That part hasn't changed. Autonomous technology might make it a little more volatile, a little more unpredictable and maybe a little scarier. It only makes sense to have a way to join 'em, rather than trying to beat 'em, which may be futile.

Someday soon, we'll all be reading a news story that will provoke this reaction: "Hey, you mean they're already flying these things without pilots?"

Join the conversation.
Read others' comments and add your own.

Comments (72)

I agree, just a matter of "when", not "if". The military and aerospace companies are testing "optionally manned" helicopters and fixed-wing, I think the Marines were interested in a "mule" helicopter that could navigate itself out to a forward position, deliver 500 lbs (maybe more) of supplies and bring back injured. These things do this all by themselves, there's no "pilot" at a remote site guiding it. Doesn't the Navy's X-47 do that too - completely autonomous? (Well, it's given a mission, i.e. CAP or bombing mission, but how it carries out the orders is up to it. We see the major auto manufacturers signing on to self-driving cars within the next few years, and there's a lot more hazards on the ground! It's just a cost thing. When it's cheaper, and proven, it'll happen. It seems a lot of airline accidents occur due to basic piloting skills atrophied.

Posted by: Peter Hamilton | October 6, 2014 3:59 PM    Report this comment

When it comes to airliners, the technology has been there for a long time. I dont remember the exact date, but more than a decade ago, Boeing test flew a 737 from Seattle to Portland without the crew touching the controls. The question is liability, not just public acceptance. If the NTSB cant blame human error, pilot usually, then the airline may not be the primarily responsible party in a crash. Do Airbus, Boeing, Honeywell, Collins, etc. want to accept that?

Posted by: Jim Lo Bue | October 6, 2014 4:11 PM    Report this comment

Let the machines do the more mundane flying. Up in Class A airspace point to point. I'll be perfectly happy down in Class E flying by my wits and being the best 'stick and rudder' pilot that I can be.

Posted by: Matthew Lee | October 6, 2014 8:13 PM    Report this comment

Pilotless transport aircraft? My guess is in about fifty or hundred years. But even fifty years from now is outside of my life span so all this does not matter to me. However, October 28 is around the corner and I stand a good chance of being alive by then and it matters. Now, can we get back to calling on the FAA's secretive and questionable practice.

Posted by: Rafael Sierra | October 6, 2014 8:40 PM    Report this comment

Over the weekend, I met with a zero-time pilot-wannabee who was waxing poetic about becoming a commercial helicopter pilot and flying for the Parks Service or law enforcement. While explaining the various training requirements, I mentioned autonomous vehicles, and asked him if he was ready to invest the cost of a Harvard education in a career that - if attainable at all - might have a shelf life of a decade or less.

I'll admit that I'm a local villain in this piece. I've been predicting and advocating autonomous vehicles for a long time. Of course, nothing's a good idea until it's YOUR idea, so I'm accustomed for waiting for other people to have brilliant ideas, with which I then can agree vigorously.

Graybeards like me (literally!) have seen this before - first it was the radio operators, then the navigators, followed by the flight engineers. Early flight attendants had to be registered nurses, and many airlines once insisted that their captains had to be A&P mechanics. In the nuttiest-to-date example of regulation for the sake of regulation, the FAA now demands that UAV operators (cinematographers, in this case) need to be certificated pilots of human-piloted aircraft. I wonder who will have to possess a pilot's license in the cases of deployment of completely autonomous vehicles? The owner? The guy who charged the batteries? The gal who flips the "on" switch?

On the bright side, autonomous vehicles are going to be the savior of general aviation. For the manufacturers, maintainers, and operators. Not for the obsolescent pilots. I'll tell you this: I'm looking forward to no longer having to fight my way through sclerotic traffic caused by hoards of insipid soccer moms hauling mini-vans full of over-supervised kids to and from every imaginable activity.

But pilots who feel threatened by all of this should take heart - the FAA is absolutely allergic to anything that upsets their stone-age paradigm of indispensability. They're terrified that UAVs are a technology that actually can safely and effectively implement a strategy of "see and avoid" - even in zero-zero weather. What do you suppose a government agency will do when faced with a technology that can make it irrelevant? Conveniently, that's no mystery, because the agencies regularly issue press releases - and worse - that demonstrate their reactions. Stay tuned and pass the popcorn.


Posted by: Tom Yarsley | October 7, 2014 7:42 AM    Report this comment

After pondering the question I think that drones will not totally substitute. Remember the "paperless society" prediction? The paper communication medium has not disappear albeit rather reduced. However, I would agree to that the "drone or pilotless air transport" technology will diminished the "manual" and increase the "auto" - human pilot intervention accompanied by dogs is unavoidable.

Posted by: Rafael Sierra | October 7, 2014 8:26 AM    Report this comment

Autonomy will happen for military, and to a certain extent, law enforcement/pipeline patrol/etc. However, I don't expect autonomous commercial passenger service to ever happen. There are just too many risks involved - computer failure (this has happened in transport category planes already), navigation signal jamming/spoofing, and as some others have pointed out, manufacturer liability. And I certainly wouldn't willingly board an airliner that I knew had no human pilots up front.

On the subject of UAV operators being required to be certificated human-piloted aircraft, this too is the wrong move. I recall some research being done (though I don't recall the specifics, so I can't link to it) that showed non-pilots were quicker to pick up on and master UAV operations than certificated pilots were. The two skills are not the same, and should require different training. I do believe UAV operators should have some training and certification, however, but based on its own syllabus, not one based on human-piloted flight.

Posted by: Gary Baluha | October 7, 2014 8:33 AM    Report this comment

Excellent article. I've been involved with military and commercial aviation for nearly 40 years, and in industrial automation for around 30 years. I agree with the slowness of certification, especially with existing airliners that would have to be retrofitted at great expense.

The huge surge of airliner orders and production we are seeing today will predictably lead to a huge lull in orders in another decade. A new surge of retrofitting and all-new airliners designed around single or no-pilot operations, will be something aircraft builders will welcome.

The real obstacle to pilotless aircraft will not be the aircraft or passengers, but airports and an air traffic control system set up to handle them. Currently there are only a handful of runways set up for autoland, and those are at major airports. They are expensive to set up, maintain, and monitor.

Initially, the expense will only be justified on large aircraft, which fly longer routes. They will need not only the planned takeoff and landing runways set up for auto-flight, but alternate airports, with multiple runways to account for weather and wind. Currently, the most sophisticated autoland aircraft have tighter limits on winds for autoland than manual landing modes.

Long flights tend to fly over remote areas, which means they will need sophisticated auto-land runways built and maintained in remote places, a pretty expensive requirement.

Technology will make these systems cheaper and eventually they will be cheap enough to replace pilots, but pilots are pretty cheap today. When pilots get expensive, I expect to see a real push for pilotless airliners. I think that is coming soon, not in pilot pay but in expensive ab-initio training as we simply run out of pilots. I give it around 20 years before automated cargo planes are flying, which is part of why we are already running out of pilots. Young people have been raised on rapidly increasing technology and automation. Doubting this kind of progress is something they see as old-people naive, progress is their only sure thing. They look at airline pilots as tomorrow's buggy-whip makers, already obsolete, and they aren't flocking to flight schools to become obsolete, haven't been for many years now. 90% of the biggest flight schools in the U.S. are foreign students, young Americans have all but stopped entering.

Posted by: Sherman Kensinga | October 7, 2014 9:33 AM    Report this comment

Sherman, the reason my generation and younger aren't flocking to professional piloting jobs is the cost of training compared to the entry-level wages in the saturated US market. Why would I have spent tens of thousands of borrowed dollars on training, followed by years (if not a decade or more) flying for food-stamp-eligible wages at a flight school and then a regional airline, sharing crash-pads with six other guys in hopes of picking up a corporate or major airline job? Instead, I went to engineering school at a fraction of the cost (and no debt) and can now afford to build the airplane I want, to fly when and where I want, and be home in my own bed at my own house every night. And I don't have to deal with customers or the TSA.

It's a no-brainer from that perspective, and has nothing to do with the position becoming "obsolete".

Posted by: Bob Martin | October 7, 2014 9:50 AM    Report this comment

Bravo Bob Martin. Spread the word of your thought process, logic, career path, and successes. Getting into the pilot career is almost comical...well...okay...it is comical. You described exactly what many of my friends in the regional airlines are suffering through. Not to mention the lack of union support, continued management beat downs, and time away from loved ones. The pilot career is a relationship killer.

And you'll appreciate this as an engineer, a recent article was published about how pilots and flight attendants are exposed to a huge amount of radiation, an occupational hazard that nobody is talking about or doing anything about. Ask Boeing, Airbus, etc. if they design their planes with radiation protection.

Posted by: Amy Zucco | October 7, 2014 10:37 AM    Report this comment

I'm going to look at aviation automation from a different aspect namely a specialized aspect of information technology. I'm an IT engineer which happens to pay for my private pilot flying. I love my computer science majors who happen to be co-workers, but they're human. Not many days go by where we're not dealing with software defect reports, requirements changes (which then risk more software defects), hardware failures, etc. which all require human intervention to remedy. I can hear the public address call now aboard a totally automated and packed A380 over the central Pacific, "Is there a computer science major on-board? Our pilot suffered a buffer over-run." The human is a miracle of adaptability. A computer program is what it is. I think I'd prefer human oversight on the flight deck.

Posted by: Robert Mahoney | October 7, 2014 11:20 AM    Report this comment

Robert, Amen.

Posted by: Rafael Sierra | October 7, 2014 11:24 AM    Report this comment

To those who actually think that on-board human pilots are a reliable alternative to those unreliable computers (the examples offered always are Windows-based PCs, which is an ignorant joke), I have three words: fly-by-wire. If such computer systems fail, it doesn't matter how many humans are aboard the vehicle. The idea that humans somehow are controlling FBW flight surfaces is a popular fantasy - that ship left port long ago. Apparently, this idea that you can't rely on computers holds little sway at Boeing, Airbus, Gulfstream, etc. "Envelope protection" is the marketing department's description of the answer to the always-raging argument about "how much leeway do we allow that human pilot, to disregard the informed good judgment of the automation?" Increasingly, the answer is "none."

Posted by: Tom Yarsley | October 7, 2014 12:06 PM    Report this comment

Thomas you missed my point (which is probably my fault for not making it clearer). It's the software. Humans develop the requirements and code to them. NASA has lost very expensive spacecraft due to syntax errors which weren't detected despite extensive testing. Testing can't possibly cover every conceivable error combination (read that as unfolding accident chain). Air France is a prime example of sensor input failure which crippled the software calculations. Even though human intervention failed, at least they had the opportunity to correct the situation. Would the software have attempted the Miracle on the Hudson landing? Don't know. It depends. Automation's "good judgment" depends on the software requirements, how they're interpreted, how they're coded, and if that fails, what next? Those software architecture decisions are usually made by humans sitting around the table some of whom may or may not be having a good day. I absolutely love computers in the airplane but when my autopilot runs away I also like the off switch. If I were trying to reduce my product's life-cycle cost of ownership (i.e. operations labor) I'd push automation too, but forgive me if I don't join you on-board.

Posted by: Robert Mahoney | October 7, 2014 12:59 PM    Report this comment

To think that managing flying automation outside or inside of the cockpit without human intervention is contemptuous of the human race.

Posted by: Rafael Sierra | October 7, 2014 2:19 PM    Report this comment

Robert, the FBW example that Thomas brings up shows that there is software in the airliners that we board today that controls vital functions of the airplane (vital as in if it fails, you are dead). And there is no off switch

Air France 447 was, unfortunately, an example of terrifying incompetence of a human driver, who was "given" an airplane in stable, straight-and-level flight and deep-stalled it all the way into the ocean.

Even for the Hudson landing, one can argue that without envelope protection the control inputs that pilots were making could have caused the airplane to stall. Whether that would have changed the outcome, I do not know.

I am a software engineer and a private pilot. Software-driven airliners do make me nervous but the winds of change are blowing that way for sure.

Posted by: Karol Zadora | October 7, 2014 5:05 PM    Report this comment

i really don't care what comes of autonomous flight...it may be useful for large scale commercial operations if it ever happens...certainly not in my lifetime. Private aviation is still about going places...and flying your own aircraft...without a schedule or specific destination many times. I will never sit in a driverless car on a public street...and not a chance i would ever ride in a pilotless aircraft.

Posted by: robert peach | October 7, 2014 7:15 PM    Report this comment

Eventually autonomous flight will be here. Will the public accept it? Eventually yes. The answer is in the thought process of our youth. Each new generation sees new technology and accepts it. I am of the first generation to grow up with having a television in the house. My parents had a radio in the house (and car) as they were growing up. We accepted these bits of technology into our homes, cars and lives. Our future generations will also accept increasing automation, until there is no pilot in the cockpit. Is this good or bad? That is difficult to tell.
Essentially you don't know what you don't know. As we sit here pondering the question there is much that we don't know. In my early engineering career, I worked on the space program. We had a lot of very smart people doing all sort of brain storming to discover all that we didn't know. Then an incident or "occurrence" would happen where everyone sort of got a dope slap type of moment. Whey didn't we think of that?
Leaving the pilot on the ground and letting HAL fly the plane will show us what we don't know. Unfortunately, people will die, changes will be made and tort lawyers will get richer. That is progress. I agree that a young person today needs to really take a look into the future and see if there is gainful employment to last until the student loans are all paid. I also agree that the FAA will do its best to ensure that the changes will occur at an imperceptible pace.
That said, there is nothing like a flight in an open cockpit, or a Champ, or Cub. Even taking the old span can out to KACK on perfect day will never be replaced by sitting looking at a computer screen. Enjoy the gift of flight.

Posted by: Leo LeBoeuf | October 7, 2014 8:09 PM    Report this comment

I'm not worried about it. Self driving cars (which I do believe will happen soon) will be more of a threat to aviation than autonomous planes.

I do expect to see UAV's handle more risky flying tasks such as crop spraying (they already do this in Brazil), and the aerial photography folks might have a problem.

Posted by: Josh Johnson | October 8, 2014 6:43 AM    Report this comment

Promise this is my last attempt to make my point. I think the word I'm missing is "creativity". Today's software imbedded in aircraft systems is great at routine/known/defined tasks. The only creative solutions on the flight deck for "unusual situations" come from the computers located in the pilot's and first officer's noggins. We have the illusion of routine flights every day. You can read about some of them in the NTSB accident database. BTW I do expect future generations will have autonomous flight. Doubt I'll join the fun.

Posted by: Robert Mahoney | October 8, 2014 7:21 AM    Report this comment

We are not really talking robots and clones here, this is about artificial intelligence becoming true intelligence. As that point nears more and more humans will be replaced by intelligent machines. It won't be an airplane with a robot pilot, it will be an intelligent airplane that flies itself. Machine operators won't be replaced by robots operators but by intelligent machines. Even the software will be written by intelligent software, not by human programmers. Remember, the intelligence need not be perfect, it just needs to be a bit better than humans.

Posted by: Richard Montague | October 8, 2014 7:22 AM    Report this comment

Oops, that should be "robots and drones" not "robots and clones".

Posted by: Richard Montague | October 8, 2014 7:26 AM    Report this comment

"I will never sit in a driverless car on a public street...and not a chance i would ever ride in a pilotless aircraft."

I remember when people made that statement about passenger elevators that lacked smiling elevator operators. What eventually changed their minds? Stairs. Lots of stairs.

The primary barrier-to-participation in GA today is the requirement to obtain and maintain a pilot certificate. Autonomous vehicles obviate that requirement. Of course, in its Kafkaesque majesty, the FAA recently insisted that UAV "operators" need to be licensed pilots - of traditional human-piloted aircraft. I wonder where the FAA would station all of the not-aboard elevator operators in Otis-ville?

"Creativity" in the cockpit most often takes the form of fatal anarchy. "Hey - watch this!!!" A properly-designed autonomous control system would have handled the "miracle on the Hudson" the same way that Sully did - simply because the river was the only flat, open terrain within gliding distance of the powerless airliner. But an autonomous system could have executed the feat in zero-zero conditions in complete darkness - something that Sully would have been hard-pressed to achieve.

I know from personal experience that autonomous control systems actually are easier to code than are ones that must pay limited fealty to real-time human intervention (a.k.a. "interference"). Properly done, they service a Maslow's hierarchy of priorities. You don't have to (nor do you attempt to) code for every conceivable combination of circumstances (as several have asserted in this space). It's all about objectives (as captured in the hierarchy) and available means (resources) of achieving those objectives. Only when the available means equal zero, do you encounter an un-resolvable circumstance. At which point, all of the creative souls in the world won't be of any help.

It offends human egos in general, and pilots' egos in particular, but having the benefit of the collective wisdom of generations of pilots, captured and distilled from millions of flight hours' of flight experience, is a better deal for aircraft occupants, than is enduring the slow and often ineffective ritual of survival that we call "learning the ropes."

And none of this impending autonomy will prevent any pilot from blowing a hole in the sky in his/her treasured flivver. The ubiquity of cars didn't lead to the extinction of bicycles, nor of jogging. I enjoy flying very much. But I also understand that sometimes, people just want to get to where they're going - quickly, comfortably, and safely. Autonomous flying machines will serve that need more efficiently and more economically. That's just progress. It needn't offend anyone's sensibilities.

Posted by: Tom Yarsley | October 8, 2014 8:34 AM    Report this comment

Engineers tend to create things just because it is possible. There's no question that pilotless flight is POSSIBLE. The question that ought to be asked is "What about the variables?"

Here in the Midwest, we have lines of thunderstorms from the Canadian border to the Mexican border. How would a pilotless aircraft penetrate that line? We've already been warned about the latency (time delay) in weather depiction. I'm not about to trust my life to a non-pilot looking at weather on a TV screen from a thousand miles away--trying to decide how much avoidance should be given to a cell.

LISTEN to the dialogue between pilots and controllers today regarding weather. Clearances are issued--deviations demanded--and clearances revised based on what pilots see out the window. How many times have you received a clearance--based on what a controller observed ON THE GROUND--and had to reply "unable"--perhaps suggesting a better route?

Yes, aircraft CAN be flown from point to point--on a benign day--without passengers. Throw in the unexpected, and that model doesn't work any more.

Posted by: jim hanson | October 8, 2014 9:39 AM    Report this comment

"How would a pilotless aircraft penetrate that line?" Using the exact same previously-agreed-upon standards every time (which situationally might call for no penetration at all). Human pilots are the biggest variables in the process. Ten pilots, faced with the same information, would come up with eleven ways of addressing the situation. That's called "process variability." Eliminating variability is the key to improving quality. Autonomous control systems utilize all available information to service their hierarchy of objectives, just as a good pilot would. Unlike human pilots (who come with a wide variety of experience and levels of skill), they would do it consistently and reliably - they never would break any rule; they never would suffer from fatigue.

Posted by: Tom Yarsley | October 8, 2014 10:50 AM    Report this comment

The Concorde SST, a wonderful concept, did not survive due to cost. Making civilian or commercial pilotless transportation practical and affordable while coupling ground and airborne systems to aircraft is this century's challenge. The concept has a long evolvement, cost will determine the implementation.

Posted by: Rafael Sierra | October 8, 2014 11:02 AM    Report this comment

Welcome to Automatic Airlines. Please take your seats, make sure the tray table is in its upright position, and buckle your seat belts. Do not be concerned that this aircraft is run entirely by computers. It has been designed for your safety, our valued passengers. Nothing can go wrong...can go wrong...can go wrong...can go wrong.


Posted by: Cary Alburn | October 8, 2014 11:09 AM    Report this comment

Didn't Popular Mechanics do this same article about 60 years ago on automated cars, cars in the sky, and complex systems of highway in the sky because of the invention of the cruise control or some like that? Yeah, keep dreaming Paul. It will never happen, and this is why.

When you pose the questions, "What can we compare this to" and "We can do it, but should we do it?", set up the pros vs. cons, and risks vs. rewards breakout, you will see that that we will never have 100% pilotless commercial flight.

For one, in every risk management scenario, a human is required to fix it. They call that person an IT person in other systems similar to this scenario. Two, we cannot even convert our ground operations to fully automated systems. We can't move luggage from checkin to aircraft in an complete closed looped automated system, how can you expect it to happen fully in the air? Three, we just had some nut job take out a whole ARTCC sector, which has been down for over a month and won't be back fully functional for another month. Just imagine that major link going TU in the automated system? Yet another is the cost. Who is going to foot the bill and billions of dollars to establish the system, or convert the current system to handle the additional capabilities, and maintain it?

Once you answer one question, another "oh yeah" question arises. Yes it can be done capability wise, but should it be done capacity wise. The answer is clearly no. Sure you can move freight this way, once you set up the separate routing system, and infrastructure to handle it. Still, you have to have pilots and ATC to initially manage it.

Posted by: Glen Keen | October 8, 2014 12:04 PM    Report this comment

The author brings up an interesting point however....it is an incomplete and dishonest premise. The premise he writes the article under is that 'technology will affect the pilot career so a pilot needs to find another career as a backup.' The missing premise is....technology will affect ALL professions, not just flying. So what field would the author recommend our 24 year old student study that will be automation proof in the next 40 years. Remember, technology increases exponentially, not linearly. Soooo, in 40 years won't we have Artificial Intelligence doing what computer programmers do now? What about accountants? Doctors? Lawyers? If you think certain professional fields are immune it is my opinion you are mistaken. Google 'The Technological Singularity' or YouTube 'Humans Need Not Apply.' Or the article 'Robot Doctors' by David Ellis is also a good one. The point is, the centralization of Big Data and immediate access to an ever increasing amount of information, combined with automation, will affect all people. And it's not that far off. Heck Paul...I heard a computer can now write a news article.

Posted by: louie remigio | October 8, 2014 12:15 PM    Report this comment

"So what field would the author recommend our 24 year old student study that will be automation proof in the next 40 years."

Designing and building automation, of course. That's a softball question that's easy to crush. In 40 years time--the career span of our imaginary 24-year-old--the machines may be building themselves, but they won't be designing themselves. Or creating themselves. Or imagining what themselves can do. (Much less trying to edit the grammar of that sentence without understanding the irony.)

If you heard a computer can write a news articles, you heard wrong. We tried to hire one...

The salient point is this: if a young person is going to invest large dollars in training and education, where does a flying job rank in the return on that investment being tanked by AI and automation? Long term? Short term? More than an autoworker? Less than a dentist? Not sure I can say and I doubt it anyone else can, either.

My own career in journalism is just coming up on the 40-year mark. During that time, I have gone from Linotype machines, to cold pasted type, to scanned copy, to network edit terminals, to PCs to web publishing systems that allow me to publish an article 10 seconds after the writing of it is complete. (Or to fix it after it's published, something we could only dream of as print reporters.)

But what hasn't changed much, if at all, is the process of deciding what's a story and what isn't and gathering and creatively piecing together the bits of information that will compel a reader to read it. This has proven impermeable to automation, although the market vitality of the job has been decimated related to how readers get information in the modern world. Different problem not related to automation.

And now, 10 seconds later, you may read what spills from my keyboard.

Posted by: Paul Bertorelli | October 8, 2014 1:23 PM    Report this comment

Another thought that no one added but I didn't have room to discuss: Cost. Many people think remote flight is cheaper than having a guy in the cockpit, but it generally is not. Especially in the civil world, where permits and expensive link infrastructure require time and expense to erect and maintain.

And that's exactly why Diamond Aircraft is making nice profits on pilot-optional DA-42 sensor platforms. It's just faster and cheaper to fly the airplane into the target zone with a guy in it then going through all the BS to do it remotely. But will it always be thus? I doubt it. At some point, the curves cross and the pilotless airplane is cheaper and perhaps more effective. This could happen in the air transport world.

Futurism is a two-bladed axe. It cuts both ways. People who say something will never happen are just as likely to be wrong as people who insist it will. Remember this quote?

"Radio has no future. Heavier-than-air flying machines are impossible. X-rays will prove to be a hoax." -- William Thomson, Lord Kelvin, British scientist, 1899.

I tend to side with believers who say these things will happen, less with the doubters. The question is when. I don't know that anyone can predict that. But if I were 24 today, I think I'd plan my training to be ready for it.

Posted by: Paul Bertorelli | October 8, 2014 1:41 PM    Report this comment

The answer is in Teleportation. Forget ATC, weather, user fees, airspace and pilotless transport.

Posted by: Rafael Sierra | October 8, 2014 1:51 PM    Report this comment

Paul, if you've ever read what Huffington Post puts out as news, you would readily believe it has to be the first feeble application of artificial intelligence. There is a lot of AI creative work being done and it is getting better.

Sure, right now it is more cost effective to use manned aircraft for the DA-42 sensor platform, but it won't be too long until the intelligent airplane can do it alone, far more cost effectively. Just eliminating the need for human interface (stick and rudder, instrument and nav displays, and the space required) will lower the cost end create a more effective airplane able to accomplish the mission with no outside input.

Posted by: Richard Montague | October 8, 2014 2:26 PM    Report this comment

Yarsley--""How would a pilotless aircraft penetrate that line?" Using the exact same previously-agreed-upon standards every time (which situationally might call for no penetration at all). Human pilots are the biggest variables in the process. Ten pilots, faced with the same information, would come up with eleven ways of addressing the situation. That's called "process variability." Eliminating variability is the key to improving quality."

That might work in military resupply missions, and it might work in areas where weather is not a factor--but for most of the world, weather IS a factor. With thousand-mile long lines of weather, the reality is that pilots make decisions on where to penetrate that line--finding the holes. Private pilots have the option of "not penetrating that line"--the reality is that scheduled air commerce--whether passengers or boxes--doesn't have that option, except in case of an extreme emergency.

It is ludicrous to trust a gamer in an office 1000 miles away, rather than someone that is actually not only on the scene, but whose own butt is on the line. Yes, pilots often make bad decisions, but you have to admit that they have a vested interest in safety. You castigate the decision-making of pilots when it comes to weather, yet you trust another INDIVIDUAL (and a non-pilot at that) to make life and death decisions when that individual has nothing on the line. Even worse, you would trust a MACHINE to make these life and death decisions? If that concept worked, we would have errorless weather forecasts by computer model--and weather forecasting isn't exactly reknown for its accuracy.

Yes, it is technically possible to fly an airplane between two destinations--but being able to think, adapt, and change still requires a human.

LOOK at what we've become--willing to rely on automation or OTHERS (including government) to keep us safe, rather than thinking and acting on our own best interests. Not for me, thank you!

Posted by: jim hanson | October 8, 2014 2:37 PM    Report this comment

Huffington and a few other pubs actually are using an app from Narrative. It's capable of simple story writing, in a rote, menu-choice sort of way. The president of Narrative said he though the app could win a Pulitzer by 2017. That I'll believe when I see it, although it might eventually.

I spoke with another company about this and said we might be interested in a such a thing. The answer was it couldn't do what we do. Which is? Know the sources to call, call them, interview, write, edit, correct.

Posted by: Paul Bertorelli | October 8, 2014 2:38 PM    Report this comment

We'll that's an interesting point Paul, but all people can't all be designing and building autonomous machines. My original point is this...at some point in the future AI will surpass human intelligence. We stopped evolving...but the machines don't have those limits. They increase...at an ever increasing rate...so the paradigm in 20 years may not leave room for saying such a craft is 'impervious to automation.' Big Data, Connectivity, the Internet of Everything and automation is going to change much more than certain professions. And if AI is smarter than people, coming up with creative designs 'ideas' doesn't seem far out there either. I once heard someone say a Replicator is the holy grail of technology(think advanced 3D printing). A future in which humans don't develop 'careers' (a recent invention...couple centuries maybe?) so they can pay for food. and energy. Maybe the human era ends and technology becomes the next stewards of the world. It really is an unknown and uncertain issue at this point.

Posted by: louie remigio | October 8, 2014 7:56 PM    Report this comment

"We stopped evolving...but the machines don't have those limits. "

Well that's news. Was there a particular date, a remarkable achievement or some notable high point when human evolution ceased? So as a species, we're done?

Posted by: Paul Bertorelli | October 9, 2014 5:02 AM    Report this comment

Jim, the plane of the future is not going to be flown by a gamer a thousand miles away, it will be an intelligent machine in command of itself. It will make intelligent decisions in regarding weather and flight decisions, no human intervention needed.

Posted by: Richard Montague | October 9, 2014 7:30 AM    Report this comment


Your insistence on repeatedly citing remotely-piloted vehicles (as contrasted with true autonomous vehicles) shows that you have completely missed or ignored my points about autonomous vehicles.

While there can be some advantages to locating pilots in some bunker thousands of miles from the vehicle, they are few and specific. For cargo-only operations and for some military operations, there's an advantage to not having to loft the environmental apparatus that sustains on-board human life. But reliance upon remote operators introduces reduced situational awareness, and it adds counterproductive latency. As you pointed out, it eliminates that portion of accountability that comes from having skin in the game, although in a military mission that sometimes can be a good thing (the pilotless vehicle is expendable).

True autonomous vehicles are not dependent upon remotely-located pilots - human or otherwise. That's exactly the point. They have full authority, and thus can do anything and everything that human pilots can do (like deal with weather). But they do it better. They never break rules; they never fatigue; they never get "creative" (i.e. become test pilots); they can possess the distilled knowledge and judgment of thousands of experienced pilots, right from the day that they are manufactured - there's no 20-year learning curve to overcome with each new "installation."

You and I are talking about two completely different paradigms. Let's try not to conflate the two - purposefully or otherwise.

Posted by: Tom Yarsley | October 9, 2014 9:25 AM    Report this comment

Jim Hanson: In case you've not notice, the ball is on your court!

Posted by: Rafael Sierra | October 9, 2014 11:14 AM    Report this comment

Yarsley--"Your insistence on repeatedly citing remotely-piloted vehicles (as contrasted with true autonomous vehicles) shows that you have completely missed or ignored my points about autonomous vehicles."

And I would argue that you've completely missed my point on the inability of RPVs or machines to see what's actually happening outside. Weather is dynamic--a thunderstorm off the arrival or departure end of the runway may be flyable one minute, but not the next.

I didn't limit my observations to only remotely piloted vehicles. NEITHER the RPV or "artificial intelligence" has the capability of making on-the-spot decisions.

Spend some time flying EVERY DAY--in small airplanes and in jets--doing cross-countries (and I mean ACROSS the COUNTRY--and the remote parts of the world--and you will realize that there is sometimes no substitute for a human being on site to make decisions. Yes, sometimes those decisions will be wrong--but more often than not, they will be based on the best information available--and regarding weather, that information is not now or likely will not be available remotely or automatically.

I've already conceded that it is POSSIBLE to fly an airplane between two points--that isn't that difficult. The weakness of RPVs or "artificial intelligence" is GATHERING that weather information. Much has been made of not using XM weather for making critical weather decisions because of the latency issue--that's child's play compared to actually having to negotiate weather deviations due to weather that changes by the SECOND. ATC has access to the latest weather, but LISTEN to the negotiations as ATC tries to route pilots in weather--it's the pilots that ask for re-routes and deviations. No pro pilot that I know would use onboard uplinked weather for penetrating a line of thunderstorms--but you would bet your life on it? Get some weather experience, then let's have this discussion again.

On the subject of automation, would you accept automated ATC? After all, the savings are as much or more than automated flight, and the problem-solving is far easier--just keep the targets from merging. The tools available are far easier to use as well--speed restrictions, holding, vectors, altitude changes. Let's do that before we take people away from actually being on site. On the other hand, there's that pesky weather thing again--how would you assure that ATC didn't route you through an area of thunderstorms? After all, it happens all the time TODAY--but pilots and controllers make up for the shortcomings of the machines.

Lest you consider me a Luddite in opposing technology, let me assure you that is not the case. I use uplinked weather, collision avoidance, GPS LPV approaches, radar, stormscope, and terrain warning every day. Perhaps the best analogy for you is autopilot use--though I regularly hand-fly approaches, when the chips are down, I always use the autopilot--after all, any dummy can fly an ILS after only a few hours of training--a PILOT is better suited to monitoring, thinking ahead, and assuring that the automation does what is expected.

Flying cargo drones in good weather? No problem--but trust my life to something that doesn't have immediate and direct knowledge of what is going on in close proximity to the airplane? Not in my lifetime.

Posted by: jim hanson | October 9, 2014 1:02 PM    Report this comment

Artificial intelligence does not have to be perfect, it just has to be better than the human it replaces. An autonomous airplane will see and respond to weather itself.

Posted by: Richard Montague | October 9, 2014 1:36 PM    Report this comment

Jim: Commendable rhetoric, apparently based on experience. New starts stand a chance in commercial aviation.

Posted by: Rafael Sierra | October 9, 2014 1:46 PM    Report this comment

Equipment fails, software has glitches. Don't think so; just look at the space program, manned and unmanned. The automation was extensively vetted, each component (millions of them) traceable to their origin, each lot tested to exhaustion, every electronic box subject to vibration, thermal, electrical, EMI, EMP, radiation and any other test imaginable to discover the infant mortalities. Software was reviewed and tested by multiple teams of engineers and programmers.

All this testing costs an exorbitant of money. Still we had failures. Somewhere in the mix we have to decide the cost benefit of autonomous flight vs. manned flight. Are you going to have dual, triple, quad, or dual dual redundant systems? How extensive will the redundancy cross strapping have to be? How will this effect weight, reliability and cost? What about initial acquisition cost, maintainability, operational cycle etc.? These are questions without simple answers.

We are going that way however, if we do it wrong, kill a lot of people and damage property, then public perception and sentiment may greatly retard autonomous flight progress. Who will foot the bill for the development and implementation? Who will develop the operating, build and testing standards?

At one time NASA filled the role of directing the space program. That was the days of Von Braun and many engineers and scientists. Now NASA is loaded with too many political hacks and lawyers. There are many questions that stand in the way of progress while waiting for answers.

Posted by: Leo LeBoeuf | October 9, 2014 2:51 PM    Report this comment

"Artificial intelligence does not have to be perfect, it just has to be better than the human it replaces. An autonomous airplane will see and respond to weather itself."

People have been trying to predict the weather for hundreds of years. Despite all of the information now available, despite all of the automated systems, they don't do a very good job of predicting AREA weather--let alone the actions of a single cell.

Let us know when you've figured out how to do that, and after you have made a couple of years of test flights. Right now, the technology doesn't exist--but because the FAA has now declared the ADS-B system operational, maybe you could get some of the unemployed guys that developed that technology to help perfect the system. They could do for you what they've done for the rest of us with ADS-B.

Posted by: jim hanson | October 9, 2014 2:56 PM    Report this comment

Regarding public perception and sentiment after the first fatal "accident" of an autonomous flying machine:

We don't have a way to estimate this reaction, as this is new territory. However, we can compare it to some contemporary events such as the runaway throttle incidents that were associated with certain Audi and Toyota automobiles. In these events, the vehicles were reported to take off uncontrollably at high speed resulting in fatal crashes. In one event, a 911 recording exists of a terrified father as he attempts to save his family from the crash.

These incidents were devastating to the public confidence in these machines, and today they still leave a cloud of suspicion in the public mind. Now, just imagine that scenario happening onboard an automated B-757 and what the public reaction will be like. For those old enough to remember, look up the history of the Lockheed Electra; I predict it would be an order of magnitude greater because the expectation of safety is much higher now than in 1959.

Posted by: A Richie | October 9, 2014 4:14 PM    Report this comment


"On the subject of automation, would you accept automated ATC?" If done correctly, I absolutely would. But if done correctly, there would be very little of it, because airborne vehicles would negotiate their own airspace conflicts - without "help" from a central ATC authority. No need for great fear - that paradigm is anathema to the FAA. That's ionic, because the FAA has insufficient capacity to disavow their "see and avoid" paradigm that's the rule for all flight in visual conditions (even under IFR). Well-designed autonomous vehicles could "see" and avoid each other - even in zero-zero conditions.

"...and you will realize that there is sometimes no substitute for a human being on site to make decisions." Politely, Jim, that's your uninformed opinion. I've spent a great deal of my career designing machines that reliably do exactly that, every day. If I can do it successfully, it obviously can't be very difficult.

"...you've completely missed my point on the inability of RPVs or machines to see what's actually happening outside." Jim, you need to get past the human limitations of "seeing." Machines can be designed to sense phenomena that are both within and beyond the abilities of humans' sensory limitations. Sully could see the Hudson river in daylight VFR conditions. An autonomous control system could sense the river in zero-zero total darkness in a snowstorm. I'm not dissing Sully; I'm pointing out that once you get past anthropomorphic ego, you can take advantage of the peculiar characteristics of well-designed machines.

When man first put an engine in a boat, he realized that figuring out how to connect the engine to all of those oars would be a waste of time. Industrial robots don't look very much like humans, for very good reason: their designers make no efforts to replicate human methods of doing their assigned tasks. So too, properly-designed autonomous control systems do not attempt to do things using human methods - nor are they limited by human abilities. An autonomous flight control system is NOT an autopilot on steroids. It's a completely different thing, precisely because it's designed to do its job without allowing for any human intervention. Authority and accountability as complementary features. It's paradigmatic.


Posted by: Tom Yarsley | October 9, 2014 4:25 PM    Report this comment

I just returned from a flight--the weather was reported "Clear" by the AWOS. The REALITY was that the ceiling was 3000' broken to overcast. When the FAA can't even get an Automated Weather Station to report correctly--a device firmly affixed to the Earth--not moving in 3 dimensions--I see little hope for accurately depicting weather that you are closing upon at 500 mph.

You may design machines that perform automated tasks--and there is nothing wrong with that--I've already acknowledged that moving an aircraft from one destination to another is possible. I question whether a machine can be developed that reads changing weather--when you can't even get accurate weather to feed TO the machine.

For years, trolley cars and railroads have been CAPABLE of running autonomously as well--but about the only examples we see of them is the shuttle from one airport terminal to another--a use that does NOT involve variables like weather, different flight conditions, etc. The shuttles all run the same direction--on the same tracks--yet we don't see the technology used in the wider sense. Railroads, subways, and trolleys still have human directors aboard--even though there are far fewer variables to account for than aircraft. Why do you suppose they have those operators aboard? Is it to solve problems? Is it because the public is unaccepting of completely automated trains?

Once again--I utilize all of the technology available that makes sense. If I had an aircraft that had autoland capability, I would use it. (Except the space shuttle--even though it has full autoland capabilities, I don't think there was ONE automated landing. HMMMMM.)

There are some things that are invented "just because we can". Anybody older than 50 recalls the inventions of one Rube Goldberg--the name has become synonymous with elaborate and unnecessarily complicated solutions to simple problems.

Not to get too political here, but it has been said that "Some people view life as it IS--others view life as they would LIKE it to be." Aviation is about the ability to make decisions--to control the machine. I don't know very many people in aviation that advocate for MORE government control (after all, we have seen the worst of the government regulators--people who believe they know better than those that actually participate in the activity--agencies with bloated budgets--people that feel they are above those they regulate--people that have little knowledge of the industry they purport to regulate. I've never met a pilot before that was in favor of MORE government control, and LESS by the pilot. I'll make my own decisions, thank you--and decline the government offer to "help". I think most pilots (and passengers!) would agree.

Posted by: jim hanson | October 10, 2014 2:57 PM    Report this comment

"Aviation is about the ability to make decisions--to control the machine." Really? A lot of people would say "aviation is about safely and expeditiously getting from one place to another." That's it. No romance; no adventure; no... pilots. Just transportation.

People like that are not a threat to your freedom to fly as you like. Autonomous vehicles are not a threat to your freedom to fly as you like.

You argue that machines are less-capable of dealing with [insert situation]. But you define the machines' inferior capabilities in human terms. That's a fundamental flaw in your logic.

Remember, ceiling and RVR make no difference to an autonomous vehicle - because it doesn't rely on human vision to get its job done. It cares not a whit about daylight or darkness; sunshine or fog. Those are human limitations. To a machine, a nighttime zero-zero landing is identical to a daylight CAVU landing. Existing rules were created to constrain operations in that human context. Take the humans (and their limitations) out of the equation, and you can craft rules and procedures that take advantage of the non-human capabilities of machines.

Posted by: Tom Yarsley | October 11, 2014 11:19 AM    Report this comment

As Paul so facturally (reality!) stated in the next to last pargaraph; it's NOT the "pilotless aircraft" thing that's the biggest threat to aviation in general, but the lack of FUTURE financial security offered to the young. Rafy, Amy and Bob Martin) get it - most here are still living in denial or bias to the "profession" do to their being in it.

Frankly, like those wonderful Big Bands of 65-75+ years ago - aviation (careers) will succumb to the "smarter" , wiser and less naive/idealistic generation who will choose $$ over "fun"!

Posted by: Rod Beck | October 12, 2014 12:28 PM    Report this comment

Yarsley--"You argue that machines are less-capable of dealing with [insert situation]. But you define the machines' inferior capabilities in human terms. That's a fundamental flaw in your logic".

HMM--you must have missed my post--"Once again--I utilize all of the technology available that makes sense. If I had an aircraft that had autoland capability, I would use it. (Except the space shuttle--even though it has full autoland capabilities, I don't think there was ONE automated landing. HMMMMM.)"

You mention your work with robotics, but I have to ask--are you instrument rated, experienced, and current? It goes to the ability to understand the problem All too often, engineers come up with solutions to problems that don't exist.

I've never argued that autoland doesn't work--quite the OPPOSITE. Only that machines not only can't deal with exceptions (like the tthunderstorm example I gave. You continue to say that it CAN be done--and done safely--yet you fail to not only explain how this artificial intelligence will be able to function--but how it is even available to obtain the information--given the lack of reliable weather information. FAA and industry cautions us not to use uplinked radar to make close-in weather decisions--so that's out. Anybody that uses on-board radar is familiar with the limitations on interpretation of that tool--it requires constant change (tilt, intensity, range) and interpretation. Just how would these work?

I don't worry too much about machines flying personal airplanes. In the autoland scenario you cite, it usually requires special certification of the airport--special certification of the equipment--dual radio altimeters, special certification of the lighting components, a VERY long runway, autobraking, a comparator for all onboard systems, and no less than THREE autopilots to accomplish that (the space shuttle used FIVE autopilots). Clearly, that expense is far beyond what the average owner--or corporate flight department--can afford--and the special runways for autoland severely limits the utility of a personal aircraft.

Robots are good at doing repetitive tasks--even doing an autoland. It isn't enough to say "someday, "they" will invent something--before autonomous flight will work, the information that it will use must be invented and proven. Until then, that's just wishful thinking--not based on anything available today or the foreseeable future.

If you REALLY believe that you have the solution to a problem--the American way has been to produce your product--throw it on the marketplace, and see if it sells. At a time when the FAA takes decades and hundreds of millions of dollars to certify any new technology, I don't think you'll have many customers for your product--or investors. If an airline DOES decide to invest in your product--they will have to sell it to their customers--people that want to RIDE on the airplane. That's the beauty of the marketplace--it sorts out what people WANT--not what somebody thinks OUGHT to happen.

Posted by: jim hanson | October 13, 2014 1:20 PM    Report this comment


Your continuing citation of equipment, rules, and procedures that are designed around human pilots and autopilots shows that you have no idea what a truly autonomous machine is. For example, talking about special certification for runway lighting is utterly silly, when the vehicle doesn't need to "see" anything. It's like OSHA issuing a complaint about the quality of workplace lighting in a shop that employs only blind people. Never fear - the FAA will do everything in its power to see to it that autonomous vehicles are pre-emptively regulated out of existence.

To answer your question about my qualifications to offer comments in this space, it may or may not be helpful for you to learn that I'm a 4-decade CFII and international-award-winning design engineer, with pioneering experience designing autonomous, mission-critical embedded control systems. I have defensible cause to believe that I understand what it takes to operate a vehicle without looking out through its windows. Truth is, an autonomous vehicle doesn't even need windows - the lack of which reduces structural complexity and weight, and thus makes pressurization and ice-mitigation easier.

I don't refer to this stuff as "artificial intelligence." Frankly, this task doesn't warrant that level of technology. And a well-designed autonomous vehicle wouldn't need to tilt or adjust the gain of a moveable radar antenna - and it wouldn't need a video display, so there'd be no need to adjust that, either. It would utilize a phased-array antenna, and would algorithmically vary the gain, with the objective of building a real-time 3-dimensional model of the air column. That's a basic element of its "sentience." But the model would not be limited to just one sensor input (radar), nor would the radar be utilized solely for weather-detection. There's lots of radar-imagable surfaces out there, including terrain and other vehicles, and the system would correlate its radar returns with ADS-B information, lidar returns, etc. And of course, it would update its "awareness model" dozens of times per second.

This stuff is cool, but it really isn't all that difficult.

Posted by: Tom Yarsley | October 13, 2014 4:37 PM    Report this comment

For all of your advocacy of pilotless vehicles, you say that "it really isn't all that difficult." If it is NOT difficult, why don't you build it? You could sell it to any number of companies--here's your chance to be a billionaire! Have you pitched your insight to the Chinese? They seem to be eagerly adopting new technologies, and have been making speculative purchases of American companies.

On the other hand, no major avionics manufacturer seems to share your enthusiasm. Have you contacted Collins Avionics about your insight? Lockheed/Grumman produces the X-47--yet they have yet to propose a passenger-carrying spinoff. Boeing and Airbus continue to produce manned aircraft instead of your pilotless version--what do they know that you don't? You would think that ONE of the major airlines would share your enthusiasm for the project--but I haven't seen them ask for a proposal. I haven't heard one corporate executive that wanted to ditch the company plane for a pilotless pod.

If you think you have a better idea--and it is as simple as you say it is--BUILD IT. The world will beat a path to your door--or more likely, will not. That's the beauty of the marketplace--it rewards good ideas, and rejects unworkable ones. It rejects THEORY--and rewards those who can produce a product that produce a product that people want and need.

To return to the question of this thread--for the foreseeable future, that means pilots on passenger-carrying aircraft. Given the need for multiple redundant systems aboard the aircraft, and the special certification of approach equipment, I don't think this will be viable in the non-scheduled and random operation of private and corporate operations. For most of us--that means that we have little to fear about a machine taking our piloting job. I'd worry more about other technologies, and even MORE about an imploding economy that makes travel itself too expensive.

Posted by: jim hanson | October 14, 2014 9:28 AM    Report this comment


Right now, I'm pretty busy looking for $10 million of Series-A financing for a spinoff from my software business - a project that's been in our "skunk works" for almost 7 years. So I won't have any time soon to enter the multi-billion-dollar aviation manufacturing business.

Your basic argument is: "if it could be done, and if somebody wanted it, somebody else would have done it by now." The entire engineering universe is populated with evidence that that's never true - regardless of what "it" is.

Posted by: Tom Yarsley | October 14, 2014 10:40 AM    Report this comment

"Your basic argument is: "if it could be done, and if somebody wanted it, somebody else would have done it by now." The entire engineering universe is populated with evidence that that's never true - regardless of what "it" is."

No--just that contrary to your claim, it isn't:
A. As simple to institute as you make it out to be, OR
B. No avionics manufacturer, airframer, or airline WANTS to do it, OR
C. The public will not accept it.\\

Or, maybe a combination of all three.

Posted by: jim hanson | October 14, 2014 3:25 PM    Report this comment

To ALL you engineer/production/technical guys (and gals?); "Engineers MAKE things; however, marketing/sales folks make THINGS happen!

Posted by: Rod Beck | October 14, 2014 8:16 PM    Report this comment


How about D: the FAA will not permit it, until the rest of the world already is using it.

Posted by: Tom Yarsley | October 14, 2014 10:44 PM    Report this comment


I'd say engineers make things possible; marketing and sales folks make the possible desirable.

Posted by: Tom Yarsley | October 14, 2014 10:46 PM    Report this comment

WRONG Mr. Yarsley: Consumers create demand; NEED and WANT; marketing/sales folks satisfy those demands by bringing goods and services to market - profitability!

EXAMPLE: "Flying Car" - what demand???????????????????????????????

Posted by: Rod Beck | October 14, 2014 10:55 PM    Report this comment


I started out in engineering and manufacturing in the early 1970s. I quickly learned two things:

1. "If you build a better mousetrap, the world will beat a path to your door" is demonstrable nonsense. The best mousetraps need good marketing and sales.

2. Selling vaporware is a great way to lose customers' trust and/or wind up in jail.

The only things that marketing/sales folks successfully bring to consumers is products that actually exist. You can sell all of the vaporware and unobtanium airframes you want to - you just can't deliver any. Example: warp drive. Lots of demand, but unless/until some engineer makes it possible, even revival-meeting-level demand doesn't matter at all.

Jim Bede created lots of demand - but damned few airplanes.

Posted by: Tom Yarsley | October 15, 2014 8:39 AM    Report this comment

Hi Thom; I guess my question is what came first - the chicken or the egg?

Simply, no matter how brilliant the product (Flying Car) from an engineering/production perspective, without consumer demand, it hasn't ANY market (sales) ability. And frankly, "demand" doesn't mean just because someone says I WANT one; REAL demand is when that "someone" PAY$ for the product or service.

The problem with ALL of aviation is the "build FIRST and "they" will come" premise - without ANY supporting demand evidence! Would you care for me to cite several examples of the ass backward approach to "selling"?

One might ask themselves WHY the Jap imports have overtaken the :Big Three" (Detroit)? They did the market (demand) research BEFORE the engineering production folks went to work - then PRODUCED the vehicle that consumers wanted and needed. The idiots in Detroit utilized the "WE think" we know (egos?) without thorough demand research - than produced "inferior" vehicles consumers weren't buying. I guess the "turtle: here won the race?

Posted by: Rod Beck | October 16, 2014 7:51 AM    Report this comment


First of all, no customer can take delivery of a product that doesn't exist (like warp drive). Secondly, any manufacturer can build a product that meets all of its performance requirements, yet sells very poorly.

Some products are pioneering in nature - the first pocket cell phone, etc. These products create markets where none previously have existed. Other products are derivative in nature - electric cars, flat-screen displays, etc. They most often seize share in an existing market segment.

Market research had nothing to do with Japanese success in their early days. Honda entered the U.S. car market with their model C, which was little more than a 4-wheel motorcycle. It was not what Americans wanted, but it was what Honda could build at the time. They knew full well that building and selling the diminutive C would be an educational experience for them, and they were willing to pay the tuition. The C begat the CV, which begat the CVCC, which begat the Civic. Very little advertising was done - Honda couldn't afford it. A few open-minded individuals bought these unusual (at the time) cars. Some loved them; some junked them. But Honda continued to improve them, and then (ah, the serendipity of good timing)... An asteroid struck the automotive marketplace: a 2x increase in the price of gasoline. Suddenly, the game had changed, and the traditional players with their traditional products were fur-less in a newly-cold world. Like many stressed species, they adapted. But their dominance vanished as quickly as the environment changed.

The lessons of Darwin are widely misunderstood. But they can be instructive. Ask anybody at Cirrus.

The malaise in light-GA manufacturing has two causes:

1. Demand for all light-GA airframes - new and used - is at historically low levels.

2. Demand for all light-GA airframes is vastly outpaced by the supply.

3. No manufacturer can build new product that will be cost-competitive with the existing supply of their own used products.

Introduction of evolutionary products will not affect this marketplace in any meaningful way. An asteroid will. What does an asteroid look like? Examples follow.

Outlaw leaded avgas without providing a fungible replacement for it at similar cost.

Introduce fully-autonomous passenger-and-cargo-carrying light GA aircraft.

Being a dinosaur can be a bitch.

Posted by: Tom Yarsley | October 16, 2014 9:43 AM    Report this comment

Hi Thom, Ok - let me "cut to the chase" here. your quest, (subtle) I think, is NOT to diminish the VALUE of the engineering/production professions (yours?) and their admirable contribution to the "free markets " of the USA (business/products) since WHO (your choice of names?) was born?

That said, help me out here; WHY is it then we see few top CEO's coming from engineering experience/education/backgrounds, etc, when most (fact) top CEO's/executives, managers and savy small business owner's are from the ranks of the sales and marketing areas of the business enterprise?

Perhaps because they get "it" SOLD?

Posted by: Rod Beck | October 16, 2014 8:24 PM    Report this comment

"WHY is it then we see few top CEO's coming from engineering experience/education/backgrounds, etc..."

There's a bifurcation at the CEO level. Publicly-owned companies increasingly are run by MBA types whose primary focus is the price of the company's stock. Their principal method is short-term earnings, which satisfies the objectives of short-term investors. In privately held and closely-held companies, stock price really isn't a factor, and in such companies we typically see operations people running the show.

I have great respect for effective marketing and sales folks. My basic point (quest??) is to pay homage to the obvious: the best marketing and sales folks can't sell vaporware and still stay out of jail. They need something to sell. If that's a product (as contrasted with a service), then it needs to be designed, tested, and then serially replicated for delivery to buyers.

I have great respect for the proper role of market research in the formulation of product requirements. But again, would-be-buyers' desires have to be accompanied by engineering's & manufacturing's ability to deliver (warp drive).

Everybody wants a brand new wonder-plane 101. I want a brand new wonder-plane 101; I suspect you do, too. Of course, it should cost less than $100k, carry four fat American adults and their golf gear 900 miles, and promote shinier hair and whiter teeth. Speaking strictly as a design engineer, there's no way I could tackle that project unless one of the conditions was minimum annual production of three million units per year. That's what it would take to avail the project of the required levels of automated production methodologies. Since there's no imaginable circumstance in which that condition would apply, the project as defined is a non-starter.

Now, what about my $10k "magic box" avionics product? It would require no new technology at all, and the FAA's 2020 ADS-B-equipage mandate arguably creates a captive market. Garmin and a couple other firms have the ability to produce the magic box. What's stopping them? They're largely satisfied with the status quo in the marketplace - selling a small quantity of a great variety of very expensive boxes, plus a larger quantity of a smaller variety of inexpensive (non-certified) boxes. The fly in the ointment? Certification hassles and costs. If those went away, the manufacturers would have no choice, because the marketplace for the first group of products would evaporate AND the marketplace for the second group of products would explode.

That's an example of how regulation affects ("strangles") a free-market economy. How many people would own smart phones if their designs had to be certified by the FAA? The marketing and sales folks would have nothing to sell, regardless of consumer sentiment.

I think that the magic box will happen. But everybody in this business knows that 1,000 times more such boxes will be sold into the retro-fit market, than will be sold into the OEM market. It's all up to the FAA, which frankly lacks motivation. It's a Frank Capra world!

Posted by: Tom Yarsley | October 16, 2014 11:20 PM    Report this comment

Thom, Many "points" I agree with; BUT, as havi MANY years in "business;": GA (foolish!), Real Estate (getting smart!) and Auto Retail/wholesale (smarter!), I've concluded that GA has and will simply have a "demand" problem.

That said, however,

Posted by: Rod Beck | October 18, 2014 1:53 PM    Report this comment

Thom, all the constraits (government included) aside, what ultimately matters is a DEMAND for the product or service - NO demand - NO income - NO profit?

Lets get on the same wave lenth/"problem" here; enineering folks, like yourself, seem to be preoccupied on making "someting" work; as in the "Flying Car'"; a great example?

So, hers's the REAL problem as I see it: nothing personal; is it the enineering guru's and wizards of smart that think they are the "brains (ego's?) do to the fact they were the ones who "put it together"?

A "disconnect" here possibly between engineering talent and "sales" people?

What we REALLY need is to get skilled and crafted guys (and gals) like yourself, JOINTLY with the market (research) people together and ask this very basic question: WHO is OUR be$T customer for "X" FIRST and foremost!

When that question is answered; BOTH marketing AND engineering/production teams set out to develop a "model" (protatype) with the corresponding $$$ figues (needed to be profitable) based on an increment number of "sales" of REALISTIC quotas - not some half baked overly optimsitic figure that isn't highly probable or obtainable - very unbias and objective (rational) as possible!

On the LSA and WHY it probably "failed" - stay turned for my "guest blog" which I'll be submitting to Paul/Av Web in the coming weeks!

And now, a word from our sponsor? KIDDING!!!!

Posted by: Rod Beck | October 18, 2014 2:26 PM    Report this comment

There will always be those of us who purposefully and intentionally load our bodies into an air vehicle of some sort, because we want to experience the feelings and sights that are only truly available in person. FPV (First Person Video) as model aircraft people like to call it is nothing more the watching a movie, you see the sights that the camera allows you to see but none of the feeling and sense of being there is present. So to those who think that manned aircraft should just get out of the way so FPV /Drone aircraft can fly anywhere and anytime they want so they can download all the movies they can take and post all their exploits on YouTube......sorry. It is just not the same. I will be up there myself, somehow.

Posted by: Roger Mullins | October 19, 2014 10:06 AM    Report this comment


You seem to be searching for a "Volksfluggen" (people's plane). Do me a favor, and bring back the Arc of the Covenant and a couple of unicorns, while you're at the store. ;-)

There always will be those with ample cash to but something new and fancy, whether it's $800k Cirrus 4-place FG singles, or multi-million dollar vacation homes. But mass-market goods? Not so much. The biggest engineering problem facing a would-be manufacturer of a Wonderplane is the existence of a quarter of a million inexpensive used light airplanes. A new 172 and a new Warrior can compete with each other, but neither can compete successfully with their ancestral heritage.

If Cessna was serious about selling light planes (they're not), they'd scoop up some well-cared-for used Cessnas; re-engine them (diesel, of course); new paint, windows, and interior; and two of those non-existent "magic box" instrument/avionics clusters; and offer them for sale at $250k or slightly less. Thing is - ANYONE can do that - you don't have to be Textron to get it done.

Actually, I think that YOU should do it. I might even buy one from you, if that SF-50 doesn't work out for me.

Posted by: Tom Yarsley | October 19, 2014 3:15 PM    Report this comment

Thom, I think you've misread my comment. The mass :production idea proved to be 'fatal" after WW II. Fact is this; no amount of mass production will ever (economies of scale) work for (recreational) GA. And yes; IF Cessna were serious; but wait, BUSINESS, my friend, is about low risk and HIGH$$ reward - something all of aviation doesn't offer and Cessna is done "playing in the sandbox" and getting serious where the profits are - and rest assured it's not in piston aircraft OR LSA's - turbines???

Restored 150/152's - nice IDEA; yes - BUT, will there be a demand - NO! Why?
The demand for a great and tried trainer as the150/152 line proved highly successful in years gone by - different demographic needs and wants of the 60's and 70's. Try this; put a "refurbished 150/152 along side say a new Sport Cruiser or Tecnam - now grab 20-25 peoples of ALL ages and demographics from your local mall, etc, Next, bring them to the FBO at ABC International (GA side) Airport, and ASK them WHICH aircraft, from the "look"only, if they were to LEARN to fly, which ONE would they choose. After "9 out of 10" (guess?) chose the "LSA", than ALSO tell them that do to the lesser requirements of the LSA to obtain a "pilots license (hours & cost) they just happened to pick the "better" one?

Can the "new" 150/152 OFFER what the LSA (market) should have been targeted for - the potential student (non-career) obviously not; doesn't (LSA) qualify - sorry supporters - Rubics Cube anyone?

More students =s MORE aircraft purchases, maintenance, fuel/storage, etc AND income to the FBO; hence $$ to get GA off "life support"?

More on the LSA and ineffective marketing in my "guest blog" coming soon to an AvWeb near you!

And on "investing" in ANY aviation related "idea" - sorry, to grown up now, but in lieu of the lemonade stand 60 years ago - can I make ANY money at it - if so - I'm open - otherwise bring me a few more lemons to squeeze!!

Posted by: Rod Beck | October 19, 2014 6:35 PM    Report this comment

Obviously engineering is something absolutely required in aviation, but I would have to agree with Rod in a lot of ways.

First, prior to designing and building an airplane, shouldn't all three contingencies be brought together to take an overview of the project? Shouldn't we have the sales and marketing team evaluate the realistic expectations of the product in the marketplace PRIOR to spending hundreds of millions of dollars in development?

An example - Take a look at the "flying car", what the hell is this for a market? NONE! Sure, an engineering feat and a novelty, but will it ever be produced beyond 50 copies? NOPE! Therefore, the proper way...and the way 95% of the other business world operates, is figure out how many you can sell at what price, THEN get the designers to tweak the product and let the engineers make it all work.

The problem with aviation, is the technical background and engineering approach to market study. It can be proved the engineering types seldom sell anything, but the sales and marketing IS what it is all about. Ever heard of Steve Jobs? That's right, could't pass engineering 101 but built a company around figuring out a product because he knew he could sell a ton of Apple computers.

Heard of Bill Gates? Same thing, can't program a tv remote but was a sales and marketing guy who saw the possibilities and knew how to market the product.

One last question - ever heard of a Harley Davidson motorcycle? A MARKETING company that just so happens to sell motorcycles, in fact 52% of the US market. Feature Vs. feature, specifications Vs specifications of a Japanese brand, forget it, the Harley fails! Yet, marketing and product evaluation for what will SELL has brought this company success.

I think I will rest my case!

Posted by: Michael Dempsey | October 19, 2014 8:21 PM    Report this comment


Engineering guys like me RARELY get to decide what to design and build. In big companies, we NEVER get to decide. Management huddles up with marketing and customer service, and devines what customers "want," and what they'd be willing to pay for what they want. Often, their conclusions are utter nonsense, but that's not important to what comes next.

After they've decided what the market needs, they come to us lowly engineers, and say "we need a widget that does "X" and costs "Y" to produce (the sales price doesn't matter to engineers; only the cost of manufacturing + design + tooling + certification, etc.). We ask "how many will we need to manufacture per year?" They look at us dismissively, as if to say "what difference does it make?" After we explain for the 10,000th time that production rate determines the difference between an inexpensive molded part and a costly hand-carved one, they lie to us, and then we do a feasibility study. If the study concludes "yes, it can be done," we then enter a phase wherein we craft (negotiate, really) a "requirements document." Ostensibly, this document defines the product. In reality, it's designed to freeze Marketing's request, so Engineering can design something that meets the specifications laid out in the requirements document. Hard as it may be to believe, more work goes into the feasibility study and the requirements document, than goes into all of the detailed design and certification work. If not, the (poorly/inadequately specified) project fails.

Rod: while I agree that a modern LSA may be more appealing to a flight school than a re-furbished C-150, I must point out that both of those products will lose the sale to a typical $20k C-150, 90% of the time. Most flight schools simply can't see any business value in spending 6-to-7 times the price of a used C-150 on one brand new (or like-new) bird of ANY feather.

But a refurb'd C-172? That's a defensible proposition. You can use it for training, for $200 hamburger runs; for instrument training; for sightseeing flights; etc. And frankly, Americans old enough to have flying-lesson money have gotten too plump to use cramped planes with 500-pound useful loads - regardless of the price, and regardless of how much new-car-smell you spray on the upholstery.

Michael: Harley has flirted with bankruptcy several times. They currently are facing a situation not unlike GA in the mid-1980s: their moneyed demographic is aging out of the sport; younger buyers are increasingly hard to find. Here's hoping that Harley fares better than Cessna, Piper, and Beechcraft did.

Most people think that the world is comprise of optimists and pessimists. As a career engineer, I'm a realist. Although I prefer cutting-edge work, I'll design gumball machines if it pays the bills...

Posted by: Tom Yarsley | October 20, 2014 8:05 AM    Report this comment

Thom, I'll give you credit; your one determined guy!

"After THEY decided WHAT the MARKET needs they come to US lowly engineers" - do I sense an esteem problem here?

Perhaps Jim Hanson will jump back in to this "ping pong match" with you - I need a rest!

Posted by: Rod Beck | October 20, 2014 5:27 PM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration