Much Nonsense About the Horizon Air Plane Theft

Friday evening’s theft of an airplane was almost as strange as the reactions that inevitably followed.  Let’s first talk about the facts of what happened, then move on to the ‘expert’ commentaries and opinions.  And, as a bonus on what has become a lengthy article, let’s also look at a similar but different incident that occurred just a few days after the better known Friday event.

What Actually Happened

At about 7.30pm on Friday 10 August 2018 a Horizon Air employee used a pushback tractor to move and reposition a Bombardier Q-400 twin turbo-prop engine regional commuter plane that was in a maintenance area at Seattle’s main Seattle-Tacoma (Seatac) airport.  He turned it 180° so it was facing out to the runways, uncoupled the tractor, and then boarded the plane, started the engines, taxied, and took off.

The plane had been parked at a maintenance facility at one far end of the airport grounds, immediately adjacent to the start/end of the three parallel runways at Seatac.  No-one would have noticed or thought anything about seeing the plane being turned by the tractor, nor would they have thought anything of the plane starting up and starting to taxi, it would only have been in the less than 30 seconds or so when he went off the general ramp area and onto the taxiway approach to the runway that he would have started to cause people to wonder what was happening.  As he approached the runway the tower did radio to him, mildly asking for his flight number, but he ignored that request and less than a minute later was in the air.

The middle of the three runways was clear, so he turned 90° onto the runway, applied power, and very shortly thereafter, with a very light plane (little fuel, no freight or passengers) was climbing steeply into the sky.  This article has good maps of what happened on the ground, I’ve not seen any hard copy of the route he flew, but it doesn’t really particularly matter and isn’t all that significant.  This article has some good video clips.

He then spent the next hour flying around the Puget Sound area, and was fairly quickly joined by a couple of F-15s from the Oregon Air National Guard that scrambled from Portland, and reportedly not just crashing through the sound barrier, but possibly getting to speeds of as high as 1500 mph on the short 130 mile run from PDX airport to the area south of SEA where the Q-400 was flying.

Maybe the fly-boys wanted to treat themselves to the exhilarating sensation of full thrust afterburner flight.  And this was a potentially unique situation – normally, airplanes that attract fighter ‘escorts’ are filled with innocent passengers, making for a very uncertain series of responses by the fighters.  Would a fighter really shoot down an airplane with 400 people on it?  And, if they don’t shoot the plane down, what else can they actually do to “escort” the plane?

That is a question that, so far, has happily never needed to be answered (and which most commentators prefer to avoid considering), but in this case, the empty airplane with only the illegal pilot on board would allow for a much easier decision if it seemed the guy was about to plunge his plane into a tall building or major population center.

The plane, piloted by a Richard Russell, eventually crashed into an unpopulated forested area on a tiny island, killing himself and destroying the plane.  This was most likely due to running out of fuel (he’d earlier indicated that his remaining fuel was getting alarmingly/critically low), or possibly a deliberate decision to ‘end it’ such as he had talked about in radio exchanges with the Air Traffic Control center and various pilots who were online to help him land the plane if he wished.

For much of that time he was in radio communication with the ground, and showed himself to be an amazingly likeable and friendly guy.  There was absolutely no suggestion at all that he was wanting to do anything dangerous (to others) with the plane, and he didn’t have any ‘ideological message’ that he wanted to illustrate with the stolen plane.  He was just a very average ordinary guy who ended up doing a really stupid thing and not knowing how to de-escalate the situation other than to kill himself, and by the time thoughts of trying to change the outcome arose, his fuel was almost gone.

Although not entirely focused, and at times showing an appreciation of the appalling situation he had gotten into, he was certainly about as exactly opposite the stereotypes of who would steal a plane as possible.  You can hear a good share of the radio transmissions here.

The Nonsense that Followed – The Difficulty of Flying

The headline stealing nature of this event naturally attracted headline-seekers who saw an opportunity to push their various narratives, and/or just to get some airtime and publicity.  Much of the public commentary that followed has been utter nonsense.

For example, many people who should know better have rushed to express their astonishment and puzzlement at how it was possible for a Horizon Airlines Q-400 plane to be taken by one of their employees.  It is true that Richard Russell – a 29 yr old who had been working for Horizon for 3 1/2 years as a ground handler, had no formal flight training or known experience flying airplanes.

Supporting this narrative, which has been diverted to suggest “there must be more that we don’t know”, Horizon Air’s CEO, Gary Beck, described Russell’s flying as involving “incredible maneuvers”.  He went on to say he didn’t know how Russell had achieved the experience that he did.  This suggests that Mr Beck has either lived a very sheltered life or knows very little about flying.  It has been said that Russell performed one or more barrel rolls – they were probably aileron rolls, which are one of the simplest things to do in a plane, but even if they truly were barrel rolls, there’s nothing incredible about doing that, especially in a likely circumstance where they would have been far from perfect (it is much harder to evaluate the purity and perfection of a barrel roll when watching a plane fly past, at a continually changing angle, than it is when on the plane).

On the other hand, his eventual crash was attributed by the Sheriff’s Department as a result of Russell “doing stunts in air or lack of flying skills”.  The degree of flying knowledge the local Sheriff’s Department has is, to be fair, probably close to zero, but surely then they should have been silent rather than rushing to make ridiculous comments.  No mention was made of the two most likely reasons for the crash – running out of fuel and a suicidal tendency, both of which were clearly signaled in the radio communications between Russell and the ATC.

Experts were stunned that a man who had never flown a plane could do so, and “simply starting up an engine takes a fair bit of knowledge” according to a Mr John Cox, a former pilot-union official.  The official went on to say that “how he got that knowledge is important” for investigators to find out.  Cox’s comments were backed up by “Ask the Pilot” blogger Patrick Smith who went as far as to say

Without some systems knowledge, some rudimentary flying skills, and a whole lot of luck, it’s more or less impossible [to start a plane]. The average person, if put in a Q400 cockpit and told to go flying, couldn’t get a propeller turning if his or her life depended on it, let alone take to the air.

These two statements in particular are even more specious than the others, because they come from ‘experts’.  As always, it suits the pilots to play up the “special skills” required to fly a plane, and to shelter being the concept of what an ‘average’ person can or can’t do.  Maybe the notional average person can’t start an airplane, but it does not require decades of experience and thousands of hours of flying time and tuition to learn how to do so.

In reality, starting airplane engines is a simple process, somewhat akin to starting an older car with manual choke.  In simple terms, put the engine in neutral, turn the ignition on, adjust the mixture setting, move the throttle forward, push the starter button.

In slightly more technical terms, as an assist for Mr Smith who says that although he used to fly an earlier model of the same plane family, it would take him some time and effort to do what Russell did, you turn the battery master on, select 3 batteries on, select Engine # 1, start it up, repeat for #2, and off you go.

Where would you learn how to do this?  Ummm – maybe Mr Cox prefers to pretend it doesn’t exist, but Microsoft Flight Simulator and similar programs (X-Plane is perhaps the best) walks you through all these things in a very realistic environment.  This is nothing new.  The first release of Flight Simulator was in 1977, although the first releases had very primitive graphics and functionality.  But for the last decade or two, the greatly more sophisticated modeling, and the realistic graphic quality (it can now be used on huge 4k monitors with very fast refresh rates giving smooth and photo-realistic motion) of Flight Simulator has been as close to the real thing as you can get short of spending tens of millions of dollars on a full-motion simulator.

Flight Simulator is so good that pilots themselves use it, and not just to study the basics of flight, but to polish up advanced flying skills.  For example, this book, on a pilot’s website, talks about how Flight Simulator isn’t just used for fun flying experiences, it can also be used by real pilots to train for advanced certifications.

And, it gets even easier than this for non-pilot ‘average’ people.  You don’t even to have to buy a copy of Flight Simulator, load it on your computer, buy a joystick, etc, and learn the FS controls.  You can simply go to YouTube and watch a video of other people showing you how to do everything from sitting down in the pilot’s seat prior to the start of a flight to getting up again at the end of the flight.

There’s no point in getting into arguments about how easy or how simple this is, particularly after we’ve had such a vivid demonstration of a non-pilot doing exactly this.  Suffice it to say that some people would indeed be unable to even understand the labels on most of the controls.  Others would quickly feel at home and be able to get a plane airborne.  Landing it may well be a different matter, but that’s not so relevant to this discussion.

Here’s the thing – it doesn’t suit the purpose or interests of any self-declared experts to concede that flying a plane is a simple skill that can be taught, even at an advanced level, about 95% through a $50 program, rather than needing to spend hundreds of dollars per hour in a ‘real’ plane.  Even the airlines themselves now acknowledge the benefit of flight simulator training, and while it is true that a real full-motion flight simulator is much more immersive and ‘real’ than sitting at one’s computer with Flight Simulator, we’re merely talking degrees of reality and value.

It is a bit like the difference between the camera on your cell phone and a dedicated full size camera.  For most of the time, and most purposes, these days a cell phone camera is sufficiently good, it is very seldom that normal people need to use a $5,000 – $50,000 camera.  It is the same with learning to fly a plane.  For most types of flight knowledge/experience, Flight Simulator or one of its spin-offs and enhanced versions provides all the skills you need.

Another source expressed amazement at how Russell had been able to program the flight computer in the plane.  Newsflash – you don’t need to program the flight management system at all if you just want to fly in manual unassisted mode, as clearly was the case on Friday night, and again, if the source had listened to the radio recording, he’d have heard Russell say that he wasn’t using the flight computer – another captain was trying to ‘help’ by suggesting he enter in a course to the flight computer so he (Russell) could then concentrate on solving the other issues confronting him.

More Nonsense – Something Must be Done About the Security Risk

Waves of other people expressed surprise and fear – how is it possible, they demanded to know, that an umpty-ump million dollar plane was sitting ‘unprotected’ on the ground so that ‘anyone’ could just waltz in and steal it.  To hear these people, this was not an extremely rare extraordinary circumstance, it was a major vulnerability that was about to flood our nation with stolen planes.

Pilots of course were quick to point to how this affirms the need for two pilots in the cockpit all the time (they’re feeling increasingly threatened by automation and the truth that one pilot is more than plenty, most of the time).  They didn’t also point out that there are a dismayingly large number of cases where planes with two pilots have been taken over by one of the two pilots, who has then suicided and crashed the plane.  Nor did they point out that by having two pilots per plane, you’re doubling the chance that one of them might be suicidal.

Other ‘industry experts’ suggested that not only must we always have two-man crews in the cockpit, but we should have every ground worker at airports shadowed by a second person at all times, too.

Other experts made the ‘safe’ suggestion that we need better screening of all personnel with access to parked planes at airports.  Apparently they believe that a once every few years screening would be able to predict a random event a year in the future that causes someone to ‘go off the deep end’ and think suicide-by-plane type thoughts.  We should point out that by all accounts Richard Russell was a well liked person and good worker, with a stable home life and generally presenting as a very low risk in every respect.

Thorough screening would involve interviewing each person, and then going out to interview friends, family, neighbors, and co-workers, plus doing a record screen to see if there were any external stress points.  This would probably require about 20 man-hours per screened employee.

Oh, according to the Seattle Times (see linked article, above) there are about 13,600 employees at Seatac airport with authorized access to airplanes.  So, 20 man hours each, per year, and if we say each screener has about 1350 productive hours a year, that means about 200 people, full-time, screening airport employees.  Let’s say a cost per federal employee of $60,000, and we’re looking at $12 million a year just to screen at Seatac.  Now multiply that by 500 or so other commercial airports.  We’ve just created a new $5+ billion a year federal program, that most probably will be 75% ineffective.  And that leaves “unprotected” all the general-aviation type airports and smaller air-fields in the country – there are 15,000+ airports in the US altogether, of which 5,000+ have sealed runways.  So, more likely, it is now a $10 billion agency.

Hmmm – that plan sounds eerily like a description of the TSA, doesn’t it!

And, wait – before we go rushing off to create another mega-government agency, there’s another relevant consideration, too.  At least at present, it is not the responsibility of an airport to control access to the planes within its boundaries.  To put this event in legal terms, this was a case of employee theft of company property.  While it is possible the airlines and airports (and, ugh, government too) might join together to create industry-wide solutions, at present, each airline is responsible for the security of its own planes.

And let’s not lose sight of the gravity of this ‘problem’.  One stolen passenger plane in no-one seems quite sure how many decades.  Not exactly the most serious problem society has to wrestle with at present, is it.

Other people have suggested that airports should have emergency procedures in place so that fire trucks could be deployed in an emergency to block runways.  We know airports with too few fire trucks and too many/too long runways, so this would require getting some more vehicles and people to crew them, and then we wonder how long it would take to deploy the firetrucks compared to how long it would take for a stolen plane to fly off.

Update :  Actually, we have an answer to that, thanks to reader Art – the FAA requirement generally is that at least one firetruck must be able to get to the midpoint of the furtherest away runway within three minutes of a call.  However, experts note that in most cases, a runway would need to be blocked in at least two places, because blocking just at the midpoint would probably mean that unloaded planes could take off and clear the obstruction in just half the runway length.  Although it might take three minutes to get one fire truck to the mid-point, in some cases (such as Seatac) the midpoint is actually the closest point to the firestation.  We’re not sure how long it would take to get two firetrucks, one to a point one-third along the runway and the second to a point two-thirds along.  We also note this report which suggests – at least at time of its publication – many airports do not comply with this three minute requirement.

It is true that parked planes on the ground are essentially unlocked.  Our experts seem unable to comprehend or unwilling to share this simple truth.

Continuing the thought process from the preceding paragraph, and as discussed in the previous section, once you’ve gained access to the plane, you don’t need any sort of key to start a plane.  In most cases, you just flip switches, with YouTube and Flight Simulator available to walk you through every step of the process.  Taxiing a plane is no big deal, and taking off is dead simple and most of the time as easy in real life as it is in Flight Simulator – more or less full throttle (chances are the plane you’ve stolen is very lightly loaded with no passengers, bags, or freight, and probably without a full load of fuel either), optionally a bit of flap, and off you go.

The Other Recent Case and Its Implications

Early Monday morning, an employee and, in this case, skilled and trusted pilot, stole his company’s plane (a twin-engined Cessna 525 business jet) and carefully flew it from the small general aviation airport and into his house, killing himself and harming the house in the process.  His estranged wife and her son, both in the house at the time, managed to escape unharmed.

The pilot was a totally trusted and respected employee who apparently was reacting after having been charged and briefly arrested on Sunday night for domestic violence against his wife.  The company’s president described the employee as “we thoroughly trusted him beyond measure”.

The airport has no traffic control tower or other monitoring of who takes off and lands.  More details here.

This other tragedy illustrates several points.  First, this time it was not a ground crew member.  It was the company’s pilot, and it was a plane typically flown solely and only by one pilot.  What use are the suggestions of having two ground crew people keeping an eye on each other in this case?  Are we to require now that every plane, no matter how big and how small, always have two pilots?  That starts to be a big deal in something like a four or six seater plane, where now half the seats are taken up by pilots!

And what sort of airport or company security would have prevented this from happening, when the perpetrator was at the top of the command/trust pyramid?

What sort of screening system would have detected this in advance, when the trigger point happened mere hours before the pilot stole the plane, and where at every time prior to that moment, his company had nothing but the highest of praise for him?

The lesson of this tragedy, we suggest, is that there will never be a practical 100% solution to this type of (fortunately extremely rare) risk.  It is ill-advised to try and seek a solution to the last tiny smallest fraction of one percent of risk, when there are other much more pressing risks we accept in our lives, and which impact us all, every day (auto vehicle accidents in particular).

There Actually is A Simple Solution

While it might seem astonishing that most planes don’t even have a $5 ignition key/lock system, remember that commercial passenger planes in particular are parked in very secure airport spaces.  The security is preventing access to where the plane is – ie the entire airport, not in preventing access to the plane itself.  And consider also that if a determined person really wanted to steal a plane, and it did have an ignition key lock on it, don’t you think that, just like how ignition key locks can be defeated on regular cars, the same thing couldn’t be done on the plane as well?

Actually, that last statement does need some further examination.  With so much of modern airplanes now being computer-controlled, including their engines (via FADECs), it isn’t just a case of ‘hot wiring’ the ignition switch any more.  Instead, you give computer commands, either directly, or more intuitively/interactively through the traditional types of cockpit controls.  It would be possible to add a digital controller that would not authorize the airplane computer to start the engines or do anything unless it was given a particular passcode  – similar to how one can password protect/prevent unauthorized logins to your phone, tablet or computer.  This could be further augmented by requiring some type of physical device as well – ‘two factor’ authorization.

This process could be made flexible, so that some people would be given a passcode to allow them to start the engines but not advance them beyond idle, allow others to take the engines briefly up to full power (for testing purposes) before killing the engines again, and another password for enough power to taxi but not to take off.

Only the great god-like beings in their fancy uniforms, aka the pilots, would be given super-secret passwords to actually be allowed to fly the plane.  You know, the ones that must have an upper case letter, a lower case letter, a digit, and a special character, etc etc.  Of course, they’d probably write these on slips of paper and leave them clipped to the plane control columns, but at least it would be a start.

Another approach would be to require the airplane to “phone home” and ask for permission to be started whenever someone tries to start it.  In a three-way communication, the plane and the requesting person could both communicate with a central computer and see if the request to start its engines and the person making the request were both authorized or not.

In reality, there are several different approaches along these lines, all of which would greatly increase the difficulty of having planes flown away by unauthorized personnel.  They also seem likely to be more effective than trying to screen the already screened airport employees some more, and none of which would be nearly as costly as additional screening would be.

So why are none of the experts calling for this?

And a Last Thought

Believe it or not, there is one variation of ‘unauthorized people stealing planes’ that is less uncommon than you might think, but which seldom gets talked about.  This happens when people repossess planes on behalf of lenders when the plane owners default on loans.

Much like repossessing cars, repossession agents will sometimes take a stealthy approach to reclaiming the plane, for fear the owner will simply fly it away and hide it.  They’ll find a way to get access to the airfield/airport where the plane is stored, enter the plane, and simply fly it away.

One wonders what lenders would think about high-tech software/password locks on planes.  They’d probably require an override access code as part of their lending process.