Subscribe by Email

Friday, March 25, 2011

Binary Thinking

Thought IS the human element of aviation.  Pilots are now the analog component in a digital environment.  Everything else on the airplane can be automated to follow programmed values and parameters.  Human thought, or to put it another way, effective decision making, IS what keeps an airplane from becoming a ballistic projectile.  In the early days of aviation, just as it is today, it was the pilot that kept the airplane from deviating from the desired flight path.  Unfortunately today it is just as likely that the crew will cause the deviation by incorrect application of the automation.  “Garbage in, garbage out.”  Only the human brain can adapt to a changing environment by prioritizing the elements of the situation to achieve a desired outcome. As human thought, a pilot’s course of action may be arbitrary, but that’s a good thing. In fact, that’s exactly the reason we still need pilots in aviation.  Sometimes a crew must accept a safe, but undesired outcome rather than attempt a high risk , but potentially more successful outcome.  Only a human can manage the goal of safe operations within a dynamic environment.

When humans take on the characteristics of a digital device, that fail-safe paradigm begins to deteriorate.  When humans see decisions as having only two choices, versus numerous variable options, they’re exhibiting binary thought.  “Human thought” contains the elements of judgment, deliberation and prioritization.  What does it look like when, in contrast, pilots become digital or exhibit binary thinking?  Decisions are based solely on regulatory limits, procedural directives or engineering and performance data.  It is the aeronautical version of, “It’s legal, therefore it must be appropriate?”  The primary goal of any pilot or crew is a safe operation.   There are, however, no clear lines between safe and unsafe.  The only way to guarantee safety in aviation is to not fly.  Therefore, safety in the context of aviation means driving the risk to an acceptably low value. However, risk never reaches zero until the aircraft is parked at the destination.

There are two Boeing 737 hull loss accidents that make this point.  Both accidents are runway excursions. Using these accidents is in no way intended to be a critique or criticism of the pilots involved, rather a description of how over reliance on procedural directives or regulatory limits versus judgment and deliberation can significantly increase risk. These crews carried out their duties with every intention of a successful outcome.    The first occurred when a crew landed on a short runway with deteriorated braking action.  Company SOP required them to use the auto-brake system for landing even though their understanding of the system was suspect.  Also, data showed a very small margin between the predicted stopping distance and the landing runway length.  Environmental and performance data provided to the crew showed the operation to be legal, but with very little margin.  There were distractions as well as inconsistencies in the runway condition.  The end result was that the aircraft failed to stop on the runway.  The second example occurred when a crew attempted a takeoff with very strong cross-winds.  The performance data and weather reports provided to the crew showed the operation to be within limits.  There were additional runways available that would have been more aligned with the wind, however, the crew did not request nor did air traffic control offer that option.  The environmental conditions proved to be outside the capabilities of the aircraft and crew.  The aircraft was literally blown off the runway.  Contrast these two accidents with a third in which the loss of power from both engines did not prohibit the crew from achieving an acceptable outcome.  In the last scenario no computer could have been able to resolve all the immediate evaluations and prioritization as efficiently as the human brain.  Computers, processing billions of “1’s and 0’s” can make computations very quickly, but they cannot make deliberative decisions

Pilots are routinely taught that there are various environmental and performance conditions that they might encounter during the course of any flight. Standard operating procedure and regulatory directives describe the boundaries of those conditions and the requisite pilot action.  The changeover between various conditions is often presented and trained as a crisp line rather than a shaded transition area.  That’s where human thought can become binary.  For example, if the crosswind limit is 25 knots, is 24 knots safe and 26 knots unsafe?  Not all conditions are 25 knot conditions.  If the wind is less than the prescribed limit is it safe for all conditions?  Do we only need the landing distance to be shorter than the available runway to be safe?  Regulatory and performance limits must always be respected. However, what governs the operation within those boundaries?  Only humans can make those decisions.  Too often over-reliance on procedural directives can distract the crew from the ultimate goal, safe operations. 

Human thought, demonstrated by effective planning and decision-making, is required to program automation, monitor and evaluate the situation and make contemplative changes based upon the real time environment to achieve the desired outcome.  The desired outcome must always be safe operations.  Decision-making is still the province of human thought.  Only the crew can formulate a plan, evaluate its effectiveness and make tactical adjustments to ensure maintenance of a safe operation.

Friday, March 18, 2011

An Interesting Discovery

I was skiing with a friend at Winter Park Colorado this week.  We had the opportunity to talk about our professions during lunches and while riding the lift.  He is involved in discovery research while I am, of course, a professional pilot.  Our occupations have two very different objectives, his trying to establish, validate and apply new knowledge and mine trying to execute well-established procedures with very little variation in the outcome.

Ironically, our activities have the same goal.  To meet our respective goals we require no difference between an intended sequence of events and an actual sequence of events.  Both of us use standardized processes to perform the steps necessary to progress toward our destinations, a safe landing or a new and useful composition of matter.  These practices or procedures are essential in obtaining the results we want.  They have been developed with great consideration and refinement.  Even inadvertent deviation from the accepted practices and procedures will likely compromise the end result.  Whether adventitious chemical reactants on the one hand or unintended pilot actions on the other, adherence to standardized practices is imperative for both of us.

Another similarity between discovery research and aviation is that adherence to standard procedure will not ensure achievement of the desired result.  Performing experiments within the bounds of accepted practices will no more guarantee a new discovery than following SOP will guarantee a safe flight.  Rote execution of accepted methods is absolutely necessary but may often be insufficient to achieve the desired outcome.

Finally, we both use sophisticated technology to do the tasks we used to painstakingly carry out by hand.  Mass spectrometers can identify compounds much more quickly and accurately than old-fashioned combustion analysis.   Flight management computers can fly the airplane more accurately and efficiently than a pilot ever could.  Machines now perform tasks intimately involved in the outcome that we previously performed by hand.  This has had the effect of removing the human from many of the subtle elements of the process that are fundamental to a successful result.  Data is analyzed; auto flight systems are programmed with the expectation that the outcome will be the desired one.  When it is not, the confusion can range from wonder to mayhem.

What then is the guarantor of a successful outcome, manifested as a new discovery or an incident free flight?  Human thought is the nexus of both our goals.  Whether the goal is to make a new discovery or create a safe operation.  Policies, procedures and best practices all contribute to create the environment we use to pursue our individual goals, but they are not a substitute for human thought.  For a successful outcome, the desired goal must start as a concept or vision, whether entirely new or routinely accomplished, whether a physical entity or a safe operation.  Our individual goals must always be consciously pursued by applying human thought to each and every step of the process.  Interestingly, how the procedure is applied is often far more consequential than the procedure itself.  When human thought is replaced with unconscious procedural compliance or instinctive manipulation of the technology, results can be catastrophic.  Disengaging the mind from either the process or the goal can be as disastrous in the lab as it is in an airplane.

I discovered that our professions are absolutely identical in one way.  Neither our tools no matter how accurate or technologically advanced nor our procedures in whatever way soundly developed, accurately trained or routinely complied with will ever be a replacement for human thought.

Friday, March 11, 2011

The Zen of Automation

I recently heard this parable and it certainly applies the technological evolution of the airline industry.

There is a story of a young, but earnest Zen student who approached his teacher, and asked the Master, "If I work very hard and diligently, how long will it take for me to find Zen? The Master thought about this, then replied, "Ten years."  The student then said, "But what if I work very, very hard and really apply myself to learn fast -- How long then?" Replied the Master, "Well, twenty years." "But, if I really, really work at it, how long then?" asked the student. "Thirty years," replied the Master. "But, I do not understand," said the disappointed student. "At each time that I say I will work harder, you say it will take me longer. Why do you say that?" . The Master replied, "When you have one eye on the goal, you only have one eye on the path."

The goal of the airline pilot has always been to get his plane to the destination.  The path is a relationship between man and machine that achieves a safe operation.

The chronological history of the airline industry can be divided into four distinct periods.  The first is the time in the 1930’s and early 1940’s when air transportation was just getting started and the only goal was to stay in business.  In that age of infancy companies frequently went in and out of business quickly.   They tried to carry any type of revenue capitalizing on the speed of a new mode of transportation.   These young companies tried everything in an effort to just to stay in business.  Planes were unreliable and navigation was “by the seat of the pants”.  It was during this period however, that Lawrence Sperry invented the autopilot.  That one innovation would forever change the relationship between pilots and their planes. 

The next distinct ssegment of airline history is the post war period from the late 1940’s until the late 1950’s.  Airlines had found out that there was indeed money to be made carrying passengers, freight and mail.  This period is characterized by great expansion.  Airplanes were becoming a viable option to trains and the interstate highway system was still just a vision.  Airlines utilized surplus pilots and equipment after the end of WWII to expand their businesses and route structure.  Bomber engineering was used to build large passenger aircraft. Propeller driven airplanes were powered by large reciprocating engines.  They flew at moderate speeds and altitudes where weather often made for spilled coffee and the airsickness bags were used mostly for their intended purpose rather than as a “SEAT OCCUPIED” placard.

The third, so-called “Golden Age of Air Travel”, period begins in the late 1950’s with the introduction of the B-707, the first practical intercontinental jet airliner, and continues through the end of the 1970’s. The Douglas DC-10, the Boeing 747, the Concorde were all examples of trying to push the limits of air transportation.  These 2 ½ decades were about bigger, faster, higher and farther. And doing it with style. Transatlantic flight in less than 3 hours was now possible.  Stewardess uniforms were created by the most famous fashion designers of the day.  Intercontinental jets featured pubs and piano lounges where popular entertainers would mingle with the passengers.  These airborne gargantuans were powered by state of the art turbofan engines. Up front on the flight deck, however, the navigation systems hadn’t changed much from the 1950s.  Finally in the late 1960s, as a windfall of the space program, inertial navigation was available to the airline industry.  Accurate navigation was now possible when out of range of land based navigation systems.

The fourth period, the high tech age we are now in, started in the early 1980’s with the advent of Flight Management Computers (FMCs) and Electronic Flight Information Systems (EFIS). The flight instruments were no longer a shoebox sized canister filled with pieces the size of watch parts.  The data is now displayed on a video screen.  All navigation, vertical and lateral is controlled by computers referencing laser gyros.  This age will be known for all the technological improvements that have been made possible by the micro-processor.  Today the airplanes don’t go any higher, farther or faster, they just do it more reliably, more efficiently and safer.  Computer technology is now used in every part of the airline industry from online reservations to the manufacture of just about every part of the airplane and its engines.  The dramatic defining change in this period of air transportation is that the pilots no longer “fly” the airplane.  There is no longer a physical connection between pilot and airplane.  The pilot now communicates with the computer and the computer flies the plane.

A profound paradigm shift took place between the third and forth periods.  From that remarkable day in December 1903 until the early 1980’s the pilot was always in direct control of the plane.  Even if the autopilot was engaged, vertical and lateral navigation was directly controlled by the pilot.  The airplane climbed, leveled off, turned, descended and landed at the hands of the pilot.  When the industry entered this new technological age, everything changed.  The Flight Management System, once programmed, could now control the aircraft both vertical and laterally with little or no input from the pilot.  The airplane could land itself in zero/zero weather.  Many duties previously performed by the pilots were removed or replaced by automation. 

This improved technology improved efficiency, reliability and safety by giving the pilot more and better resources to manage the flight.  Human nature being what it is, there also existed the opportunity for the pilots to let the airplane fly unsupervised. Sometimes with no consequences and sometimes resulting in catastrophe.  The airlines, with direction from the FAA, recognized that these technically complex airplanes needed additional procedures to standardize their operation.  Combine a complex auto-flight management system with detailed procedures and you get a scenario where pilots see procedural compliance as the goal not the means to it.  Also some complex procedures that are seldom used require additional recurrent training to maintain proficiency.  The theory being “Follow the procedures and everything will work out.”.  Although academically correct, that logic does not work well in practice.   Effective decision-making potentially becomes replaced with procedural (non)compliance.

Soon after the techno age started, the more progressive companies saw the need to look at how these new resources as well as traditional ones could be used more effectively in a crew environment.  That was the emergence of Crew Resource Management.  Initially CRM focused more on the relationship between the crew members and less on the relationship between the crew and the airplane.  As companies operated mixed types of highly automated as well less automated airplanes, this relationship between man and machine became more strained and confused.  There is the classic exchange between the pilot who is new to the automated airplane and the one who has been flying it for some time.  The first pilot says, “What’s it doing now?” and the second pilots responds, “I don’t know, but I’ve seen this before.”  In a recent engine failure of an A380, it literally took the crew hours to acknowledge and respond to the plethora of computer generated status and warning messages in an effort to assess the condition of the airplane.

Fortunately, the most profound characteristic of this techno age of air transportation will be the increased attention given to the study of human factors in aviation.  In the case of airline pilots specifically, it really means the study of defining the roles and establishing the relationship between pilot and airplane, i.e. the relationship between human and machine.  The two have both competing and complimentary roles in an aircraft and when the lines get blurred confusion occurs.  Who is responsible for what and when.  Does the machine monitor the human or vice versa or both.  In contrast to the first three periods of air transportation when the pilot “flew” the airplane, the pilot now manages the airplane.  The pilot still has the ability to fly the plane but because they are now designed to be operated primarily with the auto flight systems engaged, it actually creates a much higher workload for the crew. 

A number of recent high profile events demonstrates that pilots must maintain proficiency actually flying the airplane.  Here is the conundrum. Today’s airline pilots are disciplined to operate the aircraft utilizing as much of the technology as possible to achieve the safest, most efficient flight, while at the same time required to maintain awareness of all flight parameters as though they had none of it.

The Zen master would say, “It is the mind not the hand that flies the plane.”

Friday, March 4, 2011

The Vacuum Salesman and the Black Box

I still remember as a newly married husband the vacuum salesman.  He asked my wife if her carpet was clean, "of course" she exclaimed.  He asked if she would indulge him and go over it one more time with her current vacuum. Afterwards, he put a new filter in the model he was selling and vacuumed the same section of rug.  It picked up a substantial amount of additional dirt.  He definitely got our attention.  What choice did we have?  Buy his machine or live with dirty carpet.

Dr. David Warren was born in 1925 in the remote Northern Territories of Australia.  To receive a better education he attended 12 years of boarding school in Sydney.  His father’s last gift to him before an untimely death in a plane crash in 1934 was a crystal radio set.  David hope to pursue radio telephony and electronics, however, the war efforts were inconsistent with “radio hams” and he turned to his other hobby chemistry.  Chemistry took his career path into the fuels industry, but he is more famous for his contributions in electronics.

David was involved in the accident investigations related to the mysterious crash of the world’s first jet-powered aircraft, the Comet, in 1953.  He theorized that the cause of the accident would be obtained much easier if they knew what statements, if any, about the aircraft’s malfunctions the crew might have made in their last moments. He proposed a device to record cockpit conversation.   Aviation had little interest for such a device at the time, however, David built a prototype based on a miniature wire recorder he had purchased.  Voice and sound could be recorded on this pocket-sized device.  He built two prototypes, one for voice and one for recording aircraft parameters such as speed, altitude, heading, etc.

There was little acceptance of the devices in Australia, however, a British company bought the manufacturing rights and began to produce cockpit voice recorders and flight data recorders. Dr. Warren recalled how his device got the name "Black Box". “It was called a black box because in the records of my meeting in London when it was first demonstrated and they were so keen, one of the people in the discussion afterwards said, "This is a wonderful black box."   And a black box was a gadget box.  You didn't have to understand it but it did wonderful things.”

Ironically, the first government to make the flight and voice recorders mandatory was Australia after the investigation of a Fokker F27 crash in Queensland was unable to reach a definite conclusion in the probable cause of the accident.

The United States FAA, with support from the pilots’ union, made the flight data recorders mandatory in 1964. In contrast to the flight data recorders, FDRs. there was universal opposition among pilots to the cockpit voice recorders, CVRs, This was not surprising since most of the conversation in the cockpit during low workload could be highly personal and unrelated to the flight itself.

Even though their use was common throughout the airline industry, CVRs were not mandated in the U. S. until 1978.   Rightly suspicious, the Air Line Pilots Association had finally relented to the use of CVRs, but with clear stipulations. Only 30 minutes of recording could be kept.  There had to be a means to erase the recording after every flight.  Finally, and most importantly the information was only to be used for accident investigation.  The proverbial camel's nose was inside the tent.

Today, the NTSB, with some misguided congressional support, is now strongly advocating the routine auditing of CVRs to evaluate and manage the engagement and "professionalism" of airline crews.  This would be a galactic mistake as well as being totally unnecessary.  There is currently in use a program, FOQA,  that monitors crew and  airplane performance through optical digital recorders that continually captures thousands of aircraft parameters. With  the common goal of improving safety recognized by both the pilot's union and company management this information is de-identified and used only for safety assurance not individual corrective action by the company.  There are some non-US carriers that use this type of data punitively, but they have experienced some unintended negative consequences.  FOQA data has been used effectively to address trends that indicate corrective action is needed by the pilot group as a whole. When there are individual events of concern, the union is has been successful addressing them without identifying the pilot(s) to the company. When presented with the data, the pilot(s) self corrects.

Dr. Warren passed away in July 2010.  He lived long  enough to see his invention realize it’s potential as an essential investigative tool.  However, he also watched as his device was used for purposes that he never intended.  After accident investigation the CVR is now used primarily by journalists and attorneys to support their agendas. Aviation safety is a distant third when it comes to the non-investigative use of those recordings.  Prosecution, editorial criticism, civil litigation, liability management and punitive action have become the more common fate of CVR recordings.

When I use CVR transcripts in aviation safety classes, they are profoundly effective.   I emphasize that those involved always started their flight with the intention of success.  We must value that concept in order to understand that "we are them!".  Often those involved lost their lives in the accident being studied.  If we are to respect our colleagues’ experience we must learn from, not sanctimoniously criticize it.  The FAA and airline managements must value learning WHY pilots do what they do over everything else if creating safe operations is truly their goal.  Cesar Milan, The Dog Whisperer, is able to rehabilitate even the most recalcitrant dog, not through intimidation and coercion, but through understanding who they are and what their natural tendencies are. Anyone who says humans are much more highly evolved has never tried to lose weight or keep a New Year's resolution.

Although there was no CVR transcript involved in the Minneapolis over flight incident, the FAA's emergency revocation of the crews' certificates is just another example of putting training and safety behind other agendas.  After the incident the Captain lamented that he just couldn't believe they were distracted for so long.  Wouldn't it be nice if the rest of the industry could learn from that crew not just what happened, but WHY?  Not from people who weren't merely speculating.  Unfortunately, I seriously doubt and would be very surprised if this crew ever talks publicly about the incident. Are we so naive that we think this was the first and last crew to get themselves that disengaged?  If there was ANY reason to believe that they planned to do this I would strongly favor criminal prosecution, but that is not the case.  They specifically did not "plan" for this to happen.  So what was different about this crew?  I would like to know so I could avoid the same fate.  We will explain it away by saying they were distracted or not paying attention.  So we will teach and attend workshops where we will all agree that we shouldn't become that disengaged as though the Minneapolis crew hadn't been taught the same thing.  Didn't they know they were supposed to pay attention? The answer is obviously yes, but I ask again, WHY did they let themselves get so distracted? Never underestimate the power of a personal testimony from someone who has "been there".

I will never forget the dirt that new vacuum got off our rug.  I could be wrong, but I don't think we would have bought the vacuum or remembered the sales pitch from so many years ago if the salesman had scolded or demeaned us for having a dirty carpet.