Autonomous Commercial Aircraft – Who Asked For It Technology

In the seemingly unstoppable rollout of who-asked-for-it and what’s-next technology, automated commercial airliners seems to be a hot topic these days. To this end, a just released UBS study argues that a vast savings realized from putting pilots out of work will overcome our demonstrated reluctance to not having a Sully Sullenburger, or anyone for that matter, in the cockpit when the poop hits the fan.

Fortunately, writing for Forbes Dan Reed penned a much more cogent and ordered response to this “Wall Street knows best” report than I am able to for this blog. Read it here:

Mr. Reed argues that any lower consumer costs can only get us so far up this particular mountain because there are very significant barriers to autonomous airliners, especially when one considers that in commercial air travel, the safety “bar” is set at a zero failure rate (i.e. no fatalities). These barriers include: nowhere near ready for primetime technology, as of yet undreamed of regulatory requirements and schemes, conveniently and vastly underestimated financial barriers, and a widespread and persistent lack of consumer acceptance that will likely span generations.

The technology piece is probably the lowest of these hurdles to leap, which isn’t to say it will be anything approaching easy. In high reliability industries like commercial aviation and nuclear power generation, outright adoption of disruptive technology doesn’t just happen because a tech billionaire has a wet dream. Technological change is stepwise, and even glacial, and with very good reason. These are industries where even the smallest of human errors can result in catastrophic and outsized loss of life and property. Proponents speak as if many pieces necessary for autonomous control of aircraft are on the shelf and ready to go. Tell me how Equifax just lost the most personal information of fully half the American population to hackers, yet we are somehow ready now with AI and to harden and protect the IT, control systems, spectrum and data links needed to control and back-up an autonomous commercial airliner?  There is a nice looking, slightly used CIA RQ-170 drone now owned by Iran that might tell a very different story.

Predictably, UBS envisions that the incentives to autonomous airliners will revolve around money. The UBS report concludes that lower expenses for the airlines will inevitably mean lower ticket prices for consumers, which in turn will inexorably drive demand in our relentlessly cost conscious society. Reed however, wisely touches upon the cost of R/D, certification and investments in consumer acceptance as just some of the expense categories which don’t seem to be addressed in the UBS report.

Boeing and Airbus will have to spend untold billions developing and certifying fully autonomous airliners, but to what end? Simply lowering Delta’s costs out of the goodness of their corporate hearts? No chance. Anything to do with airplanes, especially their development and certification is very, very expensive and these staggering costs will be passed along in the sticker price of the airplanes eventually delivered. Further, we aren’t just talking about the certification of a new airplane, which is already an expensive and risky undertaking (787 anyone), but equally importantly will be the necessary development of an entirely new operational and regulatory paradigm for commercial aviation, something that will need to be paid for and proven to be as safe as the current one, a mark that is currently well in excess of 99.9% safe for the major US air carriers.   Untold years of development, testing and certification await. None of this comes cheap and consumers will be expected to pay, if they can even be found to fill the seats.

Speaking of reluctant consumers, one can envision the airlines flying empty autonomous planes around until the concept is proven to be as safe as our current commercial aviation model.  Also, Boeing and Airbus will likely have trouble finding companies willing to be on the pointy end of this particular spear because the United, American and Delta’s of the world understand the existential risk of crashing even one airplane. If history is any guide though, it’s a safe bet that airplane manufacturers will do what any industry with an army of lobbyists and an expensive and untested new product facing a questionable market would do: lobby the government to provide financial, regulatory and liability assistance with the development and certification of the aircraft and to deliver tax and liability incentives to early adopters of the new technology.

Put another way, we’ll all get to pay for technology the vast majority of us don’t want, won’t be fundamentally better than what it is replacing and won’t meaningfully lower consumer prices for a long, long time, if ever.

We are living in a golden age of aviation safety. In fact, from 2010 to the present there have been zero fatalities on US passenger airlines. That is not to say however, there haven’t been accidents, errors and mistakes made by humans, pilots included, during the past seven years. Yet high reliability industries like commercial aviation have developed an incredible safety system which recognizes and accounts for the centrality of humans to the process of keeping airplanes out of the dirt. This safety system employs a many layered approach built around humans’ inherent strengths and innate weaknesses.

Put another way, it’s because of well trained and proficient pilots that commercial airplanes don’t crash more often, not despite them. As the saying goes, the pilot is always the first person at the crash scene. Self-preservation has been a prime and generational motivator when it comes to the marvel that is aviation safety. The same cannot be said about the drone operator, computer programmer or Boeing and Airbus’ CEOs – they will all go home to their families at the end of a bad day, regardless of how many mistakes they make at work.  And if a computer was controlling US Airways flight 1549 in January 2009, it wouldn’t have even registered the loss of 155 lives stemming from its inability to conceive of reacting the way Captain Sullenberger and First Officer Skiles did to save those people and themselves on that cold day.

It’s instructive to note that the US Department of Defense reports that it lost at least 140 unmanned aerial vehicles, and probably many more that we don’t know of, during the last seven years. This is an astounding accident rate.  To be sure, a few of these were lost to enemy fire, which is a risk airliners generally don’t face. However, the overwhelming majority were not. Every one of those drone operators packed up his or her gear and went home to their families after the aircraft they were flying turned into a smoking hole somewhere. Self-preservation is a powerful motivator.

It’s possible that some disruptive technology is just a waste of money. The problem, to my mind is when the money wasted is mine (in the form of tax dollars), the return on investment doesn’t justify the expense and most importantly in aviation, safety is compromised. Perhaps the time, money (private and public) and brain cells that will have to be spent on developing, testing and certifying autonomous airliners that few consumers will want to fly in could be directed towards a more noble pursuit. Let’s get back to developing airplanes that fly higher, faster, farther, cleaner, quieter and ever more safely. We should also develop systems that enable pilots to more intuitively interface with the important and necessary automation that will help them fly these future airplanes more safely and with less fatigue.

I cannot fully envision the cockpit of 100, or even 50 years from now, but I hope it will be a marvel, blending augmented reality and seamless man-machine interfaces all focused on perfect safety – and I definitely see pilots in it. Simply put, if I’m sitting in the back, then someone’s ass needs to be on the line up front.

read more

A labor market solution to the shortage of doctors and pilots.

The labor markets for doctors and pilots are surprisingly similar and are driven in large part by similar constraints:

  • Both industries are categorized as “high reliability” and are greatly regulated by the government.
  • The work is mentally challenging and physically demanding in terms of motor skills and stamina. This naturally results in a limited pool of candidates who are able to qualify for available positions.
  • Workers must be trained and certified to operate routinely and safely in a high reliability environment where even the smallest of errors can produce outsized and even catastrophic events in terms of human, property and financial costs. Training standards are very high which serves to limit the number of candidates that qualify for positions, and in fact limits the number of individuals that can succeed at training.
  • Due to the highly skilled and inherently dangerous nature of the work, new doctors and pilots can expect to work for many years in what is essentially an apprenticeship system to further develop required skills.

Given the realities of these labor markets, it would be reasonable to conclude that a lucrative career awaits those that qualify and who can succeed during their training and apprenticeship.   Why is it then that both industries are experiencing a shortage of young candidates seeking careers as doctors and pilots?

The reasons for the current and predicted shortages of pilots and doctors are easily understood from an objective economic perspective: an increasing demand for labor and reduced supply of those willing to devote their productive work to the industries. The demand side in fact is very simple to explain: both medicine and professional aviation are retiring baby boomers at a prodigious rate at the same time these industries are experiencing rapid growth.

The supply side shortage can be explained from the standpoint of young people making a rational economic decision based on a rational risk – reward analysis. Both careers are increasingly less attractive to candidates who consider the staggering personal investments of time and money required to get to a point where they can expect a level of compensation commensurate with this investment. It takes roughly a decade to fully educate and train both pilots and doctors and at least another decade, often much longer depending on economic cycles, for a pilot to reach a reasonably high level of income with a major airline. While new doctors can leave medical school with debts in excess of $350,000, it isn’t unusual for young pilots to also have educational and training debt and other commitments approaching $200,000, yet starting salaries in the aviation industry can be as low as $25,000.

Additionally, while the adverse, low paying and unpredictable working conditions for doctors during their internships and residencies has been well documented, the onerous working conditions for young pilots attempting to start a career has only recently been discussed at a broad level. The pilot apprenticeship phase demands long hours at very low pay, generally offers very little predictability, and is measured in years. Those fortunate enough to qualify for shrinking number of military aviation training positions incur obligations of up to 15 years, and are increasingly electing to pursue other career interests outside of commercial aviation, should they decide to leave the military after this obligation.

To be sure, work for employees in many industries demands long hours, low pay (especially initially) and significant unpredictability. However, most of these do not face the the inherent labor market constraints described above, or if they do, total compensation reflects this risk both in size and when it becomes available (legal and financial professionals, for example).   Simply put, young people considering careers in medicine and aviation are voting with their feet. The perceived financial and quality of life rewards won’t come soon enough and won’t be large enough to outweigh the personal investment in time and money. Risk versus reward.

Some leaders and pundits in both industries have called for a lowering of training standards as a way to create more doctors and pilots more rapidly. Others, particularly in the aviation industry, have called for a redistribution of wealth from senior pilots flying at major airlines to the junior pilots flying at regional airlines and other low paying aviation companies and/or an increase to the mandatory retirement age of 65 for commercial airline pilots. These solutions fail to addresses the fundamental economic risk-reward problem that is driving the shortages.

Solutions that force the “invisible hand” of the labor market through regulation are fundamentally inefficient and would actually serve to worsen the supply problem. No one suggests paying Silicon Valley or Wall Street executives less so the savings can be shifted to new employees. To do so, or to raise the mandatory retirement age of pilots would serve only to further disincentivise career seekers by reducing (either early or late in a career) the total value of what should be a very difficult to obtain and therefore lucrative career. Additionally, lowering senior pilots’ pay late in their careers would further exacerbate the demand problem, as many experienced pilots would elect to leave the profession before reaching retirement age. Finally, years of data in both medicine and aviation have proven that inadequate training and experience inevitably serve only put the public at increased risk. Current training standards, especially in the commercial airline industry, are data and results driven and reflect an unmatched safety culture that learns from its errors and is dedicated to driving the accident rate to zero. Training and experience is expensive, but there are no substitutes or shortcuts in high reliability industries.

Despite arguments calling for regulatory fixes, the labor market is not broken for medicine or the airline industry any more than it is for lawyers, the NBA, Wall Street or Silicon Valley. The demand problem is real and will remain so, as long as the risk to new entrants into the aviation and medical industries remains too high relative to the reward. Therefore, any genuine “fix” must address the currently out of balance risk-reward analysis that is driving talented young people to seek less risky careers that are easier to obtain, start with less debt, and have higher starting salaries which increase rapidly. Additionally, the “fix” cannot diminish the overall total value of the respective careers or cause senior and experienced professionals to leave the industries early. Said another way, money talks — pay more and young candidates will flock to these job again.

Brian Hennessy is a pilot and principal at Blue Hawk Aviation, an aviation services and advisory company based in New Orleans.

read more

Pilotless Commercial Airplanes? No Thank you.

Brian Hennessy, Blue Hawk Aviation, LLC

The crash of Germanwings flight 9525 has highlighted the need for an assessment of practices regarding pilot break, cockpit entry procedures and the balance between pilots’ medical privacy and the good of the traveling public. However, one possible response being floated, autonomously piloted commercial aircraft, is not just “not ready for prime time”, it would almost certainly be less safe than the system it replaced. In fact, the history of aviation tells us that wholly autonomous commercial aircraft would simply allow for new risks to develop and propagate, and eventually become accidents.

Several rather obvious risks come immediately to mind when considering pilotless commercial aircraft. First, if the largest banks in the world cannot secure our personal information from hackers intent on gaining access to critical IT systems and data bases, it is a certainty that an autonomous aircraft system would also be vulnerable to cyber-attack. Second, a ground station would need to exist to provide, at a minimum, an area for “control of last resort”. Therefore, how do we secure that ground station from attackers intent on commandeering an airplane, or 200 airplanes?   Finally, how should that same ground station door likewise allow for entry should a deranged ground station operator wish to destroy an aircraft and murder its passengers? This conundrum may sound distressingly familiar.

Aside from the glaring unanswered safety, technological and cost issues associated with pilotless commercial aircraft, there is an additional, if uncomfortable point which must be considered before we consider unwinding the cockpit-as-fortress policies and procedures put in place after 9/11: These critical safety procedures and policies are in place not necessarily to protect 150 people on a given airplane, but rather to protect the 3,000 people in the building the terrorists will aim it at. We know with 100% certainly what terrorist will do when allowed access to the cockpit of a commercial aircraft.

Fully autonomous aircraft? No thank you. I’ll take experienced, well trained and well rested pilots in control of the flight my family is on over computer programmers and a “UAV safety operator”, thousands of miles away and connected to them only by a paycheck and a tenuous, vulnerable radio signal.

As tragic as the fate of Germanwings flight 9525 is, the actual chances of that same event happening to any given air traveler are absurdly remote and quite frankly do not approach, even remotely, the risk needed to consider a complete top to bottom redesign of the entire commercial airline industry. Despite this fact, I have little doubt that pilots, unions, airline management and regulators will seek to mitigate this risk in a realistic way, using the proven system safety approach that has made commercial aviation one of the wonders of the modern world.

read more

The Sorry State of the Navy Reserve’s F/A-18s

This is a bit off topic for our website, but it is an issue that I near and dear to me.  The Navy Reserve’s F/A-18s are falling apart and there is no relief in sight if the Active Duty has anything to say about it – certainly no F-35s and no “gently used” Hornets either.  The site “War is Boring” captures the issues and dire situation in this piece:

read more

Ebola: what medicine can learn from Aviation Safety

Brian Hennessy, Blue Hawk Aviation, LLC

To professional airline pilots and other aviation industry professionals, the fact that our public and private health system dropped the ball by mishandling the introduction of Ebola into the United States should not come as a surprise. Nor should there be any surprise at medicine’s initial focus on placing blame on individuals (specifically nurses) rather than where it really lies: with a broken system and dysfunctional culture which does not yet recognize that human error is inevitable and therefore must be accounted for by a comprehensive safety system.  Over many decades aviation has developed just such a system; one that allows for, recognizes and “traps” human errors before they can lead to catastrophic failures.

Modern major airlines have a collective safety record that is, simply put, without peer. They belong to a unique class of industries termed “high reliability” — those that are inherently dangerous and where mishaps will likely involve large numbers of fatalities and/or injuries.  Yet the major airlines fly tens of thousands of flights per day without incident.  Consider the “natural state” of aviation: the natural thing for airplanes to do is to fly into the ground.  Gravity will take over when errors occur and are allowed to propagate unchecked.  Other high reliability industries include nuclear power generation, large scale chemical production, certain military functions, manufacturing and medicine.

Not too long ago commercial air travel was infinitely more dangerous than it is today. According to National Transportation Safety Board (NTSB) statistics, in the 1960s and 1970s commercial airplane crashes were an almost monthly occurrence, killing many hundreds of travelers per year.  By the early to mid-1980’s fatal commercial aircraft mishaps had fallen to approximately one per quarter, but were still killing over 200 people annually on average.  Flash forward to 2013 (the latest full year data available from the NTSB) and there was only one fatal accident (not even a crash) among major air carriers, resulting in only two fatalities.  When charted, this decrease in accidents and fatal mishaps has been consistent and is clearly not by chance, repeating year over year and decade over decade from the mid-1950s to the present.

Conversely, according to a ground breaking US Government study in 2000, the medical system in the United States consistently kills between 45,000 and 98,000 people annually through preventable human error. Compare this to the commercial aviation industry, where no one has died due to preventable human error on a major US airline (a “Flag” carrier) in many years.  To be sure, there are differences between aviation and medicine.  People present themselves to the medical system sick, and some sick people will die.  However, these aren’t the people represented in the appalling numbers above, which only counts those people who have died prematurely due to an otherwise preventable error made by a medical professional.  The Dallas Ebola case brings this problem into sharp focus.  Mr. Duncan may well have died of Ebola eventually, but as the response in New York has shown, he would have had a much better chance of surviving if he hadn’t been turned away on his first visit to the emergency room, despite being symptomatic for Ebola and indicating that he had visited an affected area.  This information was collected, recorded and available, but due to human error, was not used by a decision maker.  Nor would the two nurses in Dallas have gotten sick in all likelihood, had well established protocols been followed.

As the Dallas Ebola response shows us, nurses frequently get blamed for errors, but in truth doctors, the culture they operate in and the medical system in general are often much more to blame for preventable deaths. It also highlights a common problem in industries and activities lacking a robust and mature safety culture – a focus on “who” is to blame, rather than how the system failed to prevent the undesirable outcome in question.  Except in very rare cases of intentional or willful violations, aviation safety culture long ago moved away from blaming individuals, focusing instead on investigating and learning from errors, and further developing the safety systems and culture designed to trap them.

Doctors today operate in the same type of culture that commercial airline pilots did 50 years ago. Airline captains of this time frame tended to be unwilling to accept crew member input, made decisions in a vacuum, viewed themselves as the single most important “link in the chain”, were unwilling to embrace new thinking and standardization and were quick to deflect and assign blame for mistakes.  Fortunately, in commercial aviation this culture has changed completely to one that now embraces standardization, policies, procedures and technology as a way to control and trap the inevitable errors humans will make.  This change certainly did not take place overnight and in many cases significant change was prompted only after safety investigations resulting from significant loss of life.  It also requires constant vigilance by all parties involved and a 100% commitment to safe operations.  Based on over 5100 daily departures, If United Airlines, its pilots, mechanics and other employees were satisfied with a 99.9% safety record, the company would crash over 1,800 airplanes per year.

The good news is that medicine can leap forward decades by leveraging the hard and successful work of other high reliability industries, like aviation. Doctors must begin to look at the practice of medicine not wholly as an art, but rather as an inherently dangerous industry that can benefit greatly from policies, procedures, standardization training and technologies designed to account for and trap the errors that humans inevitably will make.  Only then can the medicine begin to develop a true safety culture and like commercial aviation, set the bar on preventable human error related deaths where it belongs, at zero.

read more