Western legal systems are structured to apportion blame, and laws must evolve along with the technological landscape. Locating liability and accountability must be expanded beyond the immediate physical system of human and machine to include the actions of designers and engineers who create the software of automated and autonomous systems.
How can this technology be allowed out into the community when its performance is questionable? What if its manipulated and an adverse, unintended consequence is the result? And is the push for highly automated systems coming from governments or from the competing tech companies?
As Earl Wiener stated, ‘Exotic devices create exotic problems,’ and he questioned the assumption that automation can eliminate human error. So do I. Fallible humans are writing the software and building the automated machines, so they’re predestined to fail in some way. The automation doesn’t eliminate human error, but it does create new and unexpected opportunities for exotic failures.
In April 2018, a Tesla driver in the UK was filmed by other motorists cruising down the highway while lounging in the passenger seat. This driver had put himself in the hands of his car’s autopilot function. Fortunately, local police reacted swiftly and the driver was apprehended, fined and disqualified for eighteen months. I guess he did it because the technology provided him with the tool to do it.
Can automation ever be infallibly safe?
The recent autonomous vehicle and aircraft accidents are prime examples of the new age of exotic failures that we’re racing to embrace. Societies, companies and governments must decide how much automation is worth the risk.
29.
As 2012 dawns, I’m still stuck flying the A330. The ATSB accident report into QF72 doesn’t allay my concerns — I distrust the Airbus and its fly-by-wire operation. I’m still committed to flying as a commercial pilot but I’m not enjoying myself.
I’m dreading each flight; I’m having trouble separating myself from the security of my home to go out and make a living. Even though I pack the same clothes for practically every trip, I keep leaving that task to the last minute. Hearing the rasp of the zipper as I close my luggage is like the crack of the starter’s pistol.
I’m not the Captain Kev of old. The memories of the mangled mess inside the aircraft haunt me as I do everything in my power to prevent that from happening again.
A ten-day trip from Sydney to New York in April 2012 shakes my resolve to persevere. It’s pockmarked with a variety of challenges mixed with the debilitating exposure to turbulence and a potential loss-of-control through false activation of the aircraft’s protection modes. My ability to stay in the cockpit is under threat again.
I’m approaching the four-year mark since the accident, and my situation is worsening. I decide to seek professional assistance again. I’ve avoided medication so far, but maybe that’s the next step. I contact the welfare officer in my pilots’ association to ask for his recommendations, and he gives me a name of a psychologist. I make an appointment. I’m about to enter new territory.
At the same time, I make the professional decision to stop flying again. I start to attend appointments with a psychologist recommended by our pilots’ association. He confirms that I’m psychologically unfit to fly. But there’s a catch. ‘As a psychologist I can’t prescribe medication,’ he explains. ‘Only a psychiatrist is qualified for that.’ I stop seeing this psychologist and seek another specialist.
My treating doctor refers me to a psychiatric specialist who’s already providing assessments to CASA for the annual renewal of my flying licence. This specialist recommends the introduction of medication, in the form of a selective serotonin reuptake inhibitor (SSRI), to address my mood and body response issues. SSRIs are used to assist sufferers of depression and anxiety-based disorders. They chemically alter the brain’s neurotransmitters through increasing levels of serotonin, but they’re simply described as ‘blockers’. These are powerful drugs that can have debilitating side effects, and it takes some time to find the correct dosage.
I wonder if I’ll ever return to the life I built before that flight. My Maslow’s hierarchy of needs was hit by a MOAB – Mother of All Bombs – and I’m operating in the fractured remains of the lowest two levels.
People close to me know something isn’t right, but I look whole from the outside. I turn to analogies to describe my feelings, my realties and my frustrations. But the technical failures of QF72 and the complexities of PTSD are beyond the comprehension of the majority of my intelligent and experienced friends, through no fault of their own. It’s frustrating and exhausting to try and describe what I’m feeling to someone who’s never experienced the aftermath of a traumatic event. I didn’t grasp the fallout until it happened to me. I feel there is no one I know who can help with my unique difficulties. I am compelled to inform my company that I am not okay.
*
In the navy we would always joke, ‘There’s no such thing as a free lunch.’ This cynical reality takes on new meaning for me during this period of uncertainty.
Eventually, after extensive consultations with my specialist, I confirm my intentions to return to work in September 2012.
I visit the Qantas medical department and give them more background on the past four years. They’ll provide assessments that will contribute to the reissue of my flying licence from CASA and my requalification programme. I ask their opinion, based on my body response, nightmares and disrupted sleep patterns: do I have PTSD?
They agree that I present with criteria for PTSD, and quote DSM-5.
DSM-5 . . . What’s that?
I continue with my own research, educating myself about the injury that could potentially ruin my life.
The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) is a diagnostic tool published by the American Psychiatric Association. In 2013, Version 5 was published with specific changes to the testing and diagnosis of PTSD. There’s no better organisation to quote regarding PTSD than the US Department of Veteran’s Affairs and their National Center for PTSD. The DSM-5, used by Veteran’s Affairs, provides a checklist of specific symptoms; all must be present to confirm the presence of the disorder.
I tick practically every box. I was exposed to death or serious injury. I’m continuing to relive the traumatic experience through memories, nightmares and the stimuli of flying. My sleep is adversely affected. I’m isolating myself and not enjoying activities that I did previously. I’m easily irritated, and my boundaries of patience and frustration are easily breeched. I’m hyper-vigilant and on edge when I go to work. My DSM-5 scoreboard is pretty full.
I didn’t get a manual on PTSD after the fact, so all of this is a revelation. My GP provided a list of potential outcomes but it is up to me to write a game plan specifically for my situation. The stark reality that my career longevity might be adversely affected is a bitter pill to swallow. Each year I assess my position; am I pushing it too much to stay in the air?
But I know how to ‘push it’. After all, pushing it is part of a fighter pilot’s job description. I’d had to do it countless times. I recall, on my second deployment aboard USS Enterprise, getting the opportunity to test my limits. This flight was probably my most enjoyable and also the shortest of my carrier career.
I was a certified post-maintenance test pilot, which means I was authorised to take up repaired F14s and ensure they were completely functional before being cleared for operational use. On this day, I was asked to take up a jet to perform a simple angle-of-attack calibration check.
This check simply verifies that the indicated angle of attack displayed to the pilot is correct for the aircraft’s weight and configuration. It doesn’t take long to complete and our squadron needed the aircraft to meet the day’s flying schedule requirements.
I was required to perform this routine check during the timeframe of a launch and recovery cycle. The cycle involves launching a group of aircraft first, then recovering the aircraft that launched two hours prior. At the beg
inning of a deployment, that launch and recover evolution would take about fifteen minutes; towards the end of the cruise, about eight minutes.
I wouldn’t have much time to complete my task, and it was further complicated by the amount of fuel on board. The tanks were full – 20,000 pounds full. Even with full afterburner, I would have trouble burning and dumping enough fuel to land at the maximum permissible weight in this short time.
Fighter pilots normally sweat their dwindling fuel load, but I had way too much for this flight. So I would have to improvise and make it happen.
Our squadron’s flight deck chief had pulled every string to ensure I launched off the catapult first. In no time, I was connected to the catapult and strapped in for my wild ride.
The catapult did its job and punched me off the deck. I kept the afterburners lit and accelerated to 540 knots, which was the maximum speed with external fuel tanks fitted; the ones loaded to the brim with fuel.
Safely outside of the carrier’s landing zone, I selected full afterburner and rapidly climbed to 10,000 feet to do my check. I was there in about a minute.
Meanwhile, the launch cycle was in full swing and the recovery aircraft were circling the ship like a swarm of bees returning to the hive.
I had to work fast. With speed brakes out, I quickly slowed down to landing configuration with gear and flaps out. My RIO recorded the angle of attack for the various indications and our check was successfully completed in a few minutes. Now I had to get rid of enough fuel to land.
The fuel dump system works best if the engines are at a higher power setting. It is prohibited to dump fuel with the afterburners lit, unlike the dump-and-burn an F111 can do. Someone found out the hard way that doing this in an F14 is a no-no.
I was at 10,000 feet with 8000 pounds of fuel to burn. And the landing recovery below had commenced.
I selected full power, without afterburner, and initiated fuel dumping. I corkscrewed slowly down towards the ship using maximum g turns to help the fuel dump do its job. It was agonising to watch my fuel gauges slowly decreasing through the haze of a constant +6.5 g turn.
It was taking too long and I was running late. The recovery was complete and all eyes on the flight deck were directed skyward now at my lone F14, which, with a white cloud of fuel vapour pouring out of the tail, no doubt looked like a novice skywriter.
The Enterprise was churning through the ocean, its long wake signalling to me that I better pull my finger out and get aboard. But I was still too heavy. And now I was about to join the landing pattern and had no altitude to play with.
I secured the fuel dump as I descended to 600 feet behind the ship. My only method available now to decrease my weight was the afterburner. I lit the fires and advanced the burners to full Zone 5. Manually I swept the wings back and slammed my tailhook down as my F14 accelerated like a bat out of hell towards the wake of the Big E.
For a normal recovery, an F14 would join the landing pattern at around 350 knots, extending upwind of the ship for 45 seconds before breaking into the landing pattern. This gave the pilot ample time to prepare the aircraft for landing. You don’t want to rush something as potentially dangerous as a carrier landing.
I didn’t have time for this type of normal approach; I had made the ship wait long enough. My speed increased to 540 knots quickly as I pointed the nose of my jet at the back of the ship. Through my mirrors, I could see intermittent clouds of water vapour flash over my wings. I was going fast and the air pressure was compressing the moisture in the air.
It must have looked ominous from the deck of the Enterprise as this F14 thundered towards the fantail of the ship in full afterburner, the wings enveloped in cloud. I had my aircraft on its side in 80 degrees of left bank as I crossed the wake. I increased the bank to 90 degrees and pulled to 6.5 g as I passed the stern. I kept turning hard, my nose and canopy the only parts visible as the F14 was enveloped in vapour.
Could I pull this off? I hadn’t explored this extreme edge of my capability before. Well, if you’re gonna screw up, at least look good.
I kept the wings in full aft sweep as I used the aerodynamic drag of the wing and maximum g to wash off speed. I continued to turn hard and line up for landing as I sweated my speed versus my ever-decreasing distance to the ship.
At 250 knots, I threw the landing gear down. Passing 225 knots, I selected the wing sweep to auto, allowing them to sweep forward, and selected full flap. I was still in 45 degrees of bank and rapidly approaching the wake with the engines at idle power.
Rolling out on final approach, I was still fast but also displaced further back from the ship than normal. My RIO called the ball and our approach was completed without incident.
I was buzzing with adrenalin as I made my way to our ready room. What a ride!
As it turned out, my superiors weren’t so excited. It was the Cold War, and we were operating in close proximity to the eastern islands of the Soviet Union. The day before, a Soviet attack submarine had been photographed at periscope depth in the wake behind Enterprise.
Sometimes you have to take a calculated risk and throw all your chips on the table. You’ve got to push it. Will pushing it now help me to defeat PTSD?
My research helps me to understand my frustrations, accept the emergence of my symptoms and become aware of the physiological outcomes associated with PTSD: the ‘psychobiological’ part.
The body responses I’m experiencing outside of flying, such as embarrassment and frustration through social interaction, are unmuffled. It’s as if someone has taken an iron bar to my exhaust system, knocking out all the baffles and removing all the inhibitors to my fight-or-flight chemical production. My system is being flooded with cortisol, the main stress hormone. I’m getting free-flow for the smallest stimuli.
These realities have taken time to expose themselves, and it’s apparent how easily they can remain hidden, as long as the mind is busy and distracted. My reasons for stopping to analyse my situation again are now justified and critically important to my own self-awareness. My research also confirms that the medication may help improve my quality of life and possibly preserve my career. It is a painfully slow journey of discovery.
Am I just supposed to suck it up, be a ‘tough guy’ and suffer in silence? No, I won’t remain quiet. But ultimately, it’s up to me to sink or swim.
PTSD drops a sufferer into an emotional minefield without a map; you either remain stationary and hope for the best, or try to navigate your way out. Every day, I have to figure out a new way to navigate out of danger while maintaining my shredded self-esteem and security.
*
In March 2014, Malaysian Airlines flight MH370, a Boeing 777, disappears without a trace en route to Beijing. Was it pilot suicide or an automation failure?
Soon after, in April, I’m operating QF24 from Bangkok to Sydney. The flight encounters an extensive area of thunderstorm as we cross the Australian coast near Darwin. We spend the next ninety minutes dodging weather and enduring turbulence. We’ve been cruising at a reduced speed with the seatbelt sign on for most of that time. Our passengers sleep in the darkened cabin.
As we pass abeam of Alice Springs on our southeasterly course to Sydney, the skies clear and the turbulence subsides. We reset our cruise speed, and I notice the tailwind at our altitude is almost directly behind us at 180 knots. We’re scooting along nicely now. A Thai Airways 747 is behind us, 2000 feet below and on the same track to Sydney.
I hand control over to the first officer and leave my seat to retrieve the aircraft’s technical log from its stowage, planning to make a routine entry. I place it on the glareshield and take a short break to stretch in the area behind our seats.
Abruptly, the turbulence starts again.
As I scramble back to my seat, my first officer turns on the seatbelt sign while manually reducing the speed for turbulence penetration. Not even five seconds has elapsed.
My instrument scan reveals the makings of a jet upset: my aircraft has been upset b
y the changing environmental conditions and is deviating from its stabilised flight path. My instruments reveal indicated speed at maximum, altitude 39,300 feet and increasing, vertical speed climbing rapidly, and Autopilot 1 engaged. Autopilot still engaged? It should either hold altitude or disconnect itself. As I digest this snapshot of the jet’s extreme performance profile, a further scan of the navigation display indicates a headwind has developed very rapidly.
I have to make a decision within the next second. Do I leave the autopilot engaged and let it manage the jet upset – probably propelling everyone into the ceiling as it pushes the aircraft back down to 39,000 feet – or do I disengage the autopilot and manually arrest the trajectory to prevent the aircraft from pitching down to its commanded altitude?
I disengage the autopilot to stop the climbing trajectory. My trusty first officer extends the speed brakes on his own initiative. The aircraft porpoises aggressively around 39,300 feet despite my manual intervention, abruptly climbing and descending in vertical speed and I notice on my navigation display that the Thai Airways aircraft has departed its altitude by 400 feet and is still climbing behind us.
My Airbus is in a washing machine of turbulent air as I fight with small sidestick inputs to maintain a constant nose attitude. The up-and-down shifts in vertical speed make it feel like we’re on a roller-coaster, reminding me of the first QF72 pitch-down.
It’s about 4 a.m. in Sydney, not the best body-clock time to be dealing with a jet upset on a pitch-black night. Maintaining a constant nose attitude on my display is almost impossible.
No Man's Land Page 22