Human Factors: A host of bad decisions

The Beech Baron punched out of the thousand foot ceiling at a steep angle, left-wing low, descending at a mind-boggling 16,800 feet per minute, then smacked into the ground with such force that a doorbell camera a mile away recorded the sound of the crash.

The pair of commercial pilots aboard — both employees of a Part 135 freight operation — were killed instantly.

From takeoff to crash, their flight was a mere nine minutes long. 

The Nine-Minute Flight

On that night in early January 2022, the weather could have been better. But much of commercial aviation follows the old motto of the Post Office: Neither snow, nor rain, nor heat, nor gloom of night…

The gloomy night of the flight there was an overcast ceiling at 1,000 feet AGL, an AIRMET for IFR conditions over pretty much the whole route, and a second AIRMET for low level wind shear. Moderate precipitation was falling, and the NTSB’s analysis of the weather concluded that moderate turbulence was likely present at the Baron’s 8,000 foot initial altitude assignment.

After takeoff, a TRACON controller noticed that the Baron had begun a gradual left turn off course and was slowly losing altitude. The controller queried the airplane and received a garbled radio transmission. Then the Baron disappeared from radar. No distress call was received.

The NTSB investigation would find nothing mechanically wrong with the aircraft, and the board, in its final report, declared the probable cause of the accident to be spatial disorientation.

So how did two professional pilots lose control of their flight just nine minutes after takeoff?

The Pilots

The pilot in command was 35 years old with commercial, multi, CFI, and sUAS tickets. Her total time was 1,274 hours, with 74 hours in twins, and 54 in make and model. She had only racked up 36 hours in the previous month.

Her age suggests that she was a career-change professional pilot. She was still short of the specific hour requirements for Part 135 qualification — more on that in a minute.

The copilot also held commercial, multi, CFI, and sUAS tickets. He was 55 years old with 7,697 hours and was fully qualified as a Part 135 pilot.

That said, of his time, only 115 hours was in twins, with 60 in make and model — barely more than the PIC.

He, too, had a light number of hours in the previous month and he, too, was executing a career change. He had recently retired as a sergeant in the Ohio State Highway Patrol’s Aviation Unit.

Both pilots worked for an Ohio-based Part 135 operation, which focused on carrying medical freight using a mix of twin-engine Beech Baron pistons and Lear 35A jets. Pilots hired by the outfit flew both airframes, usually starting out as combined Baron captains and Lear first officers as they built time.

On this flight, one of the company’s standard routes, an empty cargo Baron was to be flown out of the company base at Spirit of St Louis Airport (KSUS) on the outskirts of St. Louis, to Centennial Airport (KAPA) in Denver, where it would pick up cargo and return to St. Louis the same night.

The empty positioning leg could be flown under Part 91 regs, allowing new-hire pilots to build time to qualify as Part 135 pilots. The modus operandi was to pair a Part 91 new-hire with a more experienced Part 135-qualified partner pilot for the run. The Part 91 pilot was the PIC on the empty outbound time-building leg, then the Part 135 pilot would take over as PIC and fly the cargo leg back. At least that was the plan.

Before working for the Part 135 company, both pilots largely flew VFR, but both were instrument current.

The Part 91 pilot for the flight had 45 hours of simulated instrument and 21 of actual, a third of them in the Baron within the previous month. She only had 86 hours of night flight.

The copilot’s instrument time is more complicated with different reported numbers in different documents, with his company records and his logbook in disagreement. In all likelihood, he had a couple hundred hours of simulated instrument and 14 of actual, with all of the actual hours in make and model, but with only three hours within the previous month. He had logged 158 hours of night flight.

But even though the copilot was more logbook senior, both pilots had been hired by the company only three months before the crash, and were actually in the same indoctrination class.

The Part 91 pilot finished this initial company training in mid-November, about seven weeks before the crash.

The copilot completed his training at the end of November, about five weeks before the crash, running behind because of an illness that had interrupted his training.

Even though they were hired at the same time, the older male pilot qualified as a Part 135 jockey thanks to the hours he racked up in law enforcement flying.

Another pilot who worked for the company told investigators that he was “kind of surprised” to see them listed together on the schedule, as they were both “pretty new,” but that they both seemed “perfectly competent.” While not, using his words, the “dream team,” nothing stuck out to him that said they shouldn’t be flying together.

That feeling, apparently, wasn’t shared by everyone.

The Text Messages

An FAA inspector interviewed the deceased Part 91 pilot’s boyfriend, who stated that he received texts from her just before the flight, expressing concerns about the flight due to weather “and the other pilot’s skill set.” Specifically, she said her copilot “did not do a lot of IFR flights and was not very confident in his IFR abilities.”

That said, even though the copilot reportedly stated a lack of confidence in his abilities, he went.

In a written statement for the FAA, the boyfriend added that his girlfriend “was worried that the flight was dangerous and the risk was high.”

Interestingly, the boyfriend seemed to be under the impression that his girlfriend wasn’t the PIC, and he related that she didn’t feel her older copilot had enough experience to make the go/no-go decision for the weather conditions.

He said she also felt that the company would not back a no-go decision — that company policy “required” her to make the trip.

In later interviews with the NTSB, both the company’s operations director and the chief pilot told the board that one or the other of them are always available to pilots who have concerns, and claimed to have a “very much open door policy.”

Several other pilots interviewed confirmed that this was true, but when pushed, said that while that was the message from the top, the pilots themselves had personally never taken advantage.

So was this a management failing or the pilot subculture? Or was it as the low woman on the totem pole, she didn’t want to be seen as being difficult?

The Part 91 pilot’s boyfriend wrote, “She said that she wasn’t sure she would make it out of this.”

And yet, she went.

The Scheduling

After the fact, management said that they briefly considered whether the pairing of two new hires was an issue, but that both pilots had passed their training, and each met the legal requirements for their roles on the flight.

The chief pilot also said that he asked each pilot, individually, if they were OK with the assignment, and reported that neither of them expressed concerns to him — either a week before the flight, when it was scheduled, or on the night of the flight.

The director of operations said, on the pairing of the two new hires: “They both had passed their check rides, they both had completed the training profiles that we had for them.”

Additionally, he said that they had both been sent out on the line with other pilots previously, so it wasn’t like they were put together for their first-ever flight.

In not so many words, he was saying at some point you have to kick the chicks out of the nest.

Of course, the problem with this “sink or swim” approach is that sometimes people sink.

The Airplane

The airplane was a 1981 twin-engine Beech 58 Baron, 15 hours out of a 100-hour inspection. It was apparently a drafty workhorse with a total airframe time of 24,000 hours, and a door that sometimes liked to pop open in flight.

Of note, it did not have dual controls, instead having only a single “throw-over” yoke. Company officials had chosen not to install the optional dual control column.

Realistically, the copilot, regardless of his abilities — or lack thereof — was just along for the ride.

Throwing the yoke required pulling a T-handle latch on the back of the control arm, then physically moving the yoke from one side of the cockpit to the other.

Oh, right. You can only throw the yoke over at high power settings. At reduced power settings, the throttles block the movement of the control column.

Not that either pilot would have thrown the yoke over, anyway. The chief pilot told the FAA, “It is our policy to not throw over the yoke during the flight…or period. Just period.”

In fact, one of the company’s pilots told investigators that the pilots weren’t even taught how to throw the yoke, and admitted he wasn’t even clear on the procedure.

Additionally, all the steam gauge flight instruments were on the captain’s side. So not only could the copilot not control the aircraft, he really didn’t have a good view of the instruments that were displaying what the airplane he couldn’t control was doing. He might as well have been in back with the cargo, grabbing a few Zs.

Analysis & Discussion

I don’t think we need to discuss the throw-over yoke. Well, not discuss it much, anyway.

The company justified not investing in the dual control conversions for its Baron fleet because the Baron is a single-pilot airplane.

I’d be OK with that logic, but only if it operated the Barons as single-pilot airplanes.

Instead, they were manning the airplanes with two-pilot flight crews, pairing newbies with veteran pilots (or newbies with higher-hour newbies in this case) and that, in my mind, requires dual controls. Two pilots, two sets of flight controls.

For what it’s worth, in the months following the accident, the company retrofitted its entire Baron fleet with dual yokes. Better late than never, I guess, but better not late would have been even better.

The NTSB’s full determination of probable cause was: “The pilot’s loss of airplane control while flying in night instrument conditions due to spatial disorientation and the flight crew’s inability to recover from an unusual airplane attitude.”

Meh. It strikes me as unfair to say the “flight crew’s inability to recover” when only half the flight crew has flight controls.

And while I have no doubt that the crash was, indeed, caused by spatial disorientation, that was just the final straw.

I’m more interested in how it was that two professional pilots got themselves into a position where they could become spatially disoriented in the first place, especially so quickly after takeoff.

As there’s no shortage of questionable decision-making that led to that circumstance, on whose feet should we lay the blame?

Is it the “fault” of the company’s management, who sent two largely VFR pilots off into night instrument conditions together?

Or does the buck stop with the PIC who took time to express her doubts about the flight to her boyfriend, but not to her supervisors? A PIC who stated that she wasn’t sure she’d “make it out of this,” but went anyway?

Or the copilot who accepted a flight he felt unqualified for. A flight where he was the backup man, but where he didn’t even have flight controls?

The Takeaway

So the obvious takeaway is that, for low-time professional pilots, the internal pressure to get the time and experience needed to move up is tremendous.

Even though all pilots have the legal authority to say “no” to a flight — and even when they have a supportive corporate culture to do so — they may not choose to do it. I think we all sort of know that, but it’s a good reminder, if a stark one.

But there’s two other — truly alarming — aspects to this accident that are not mentioned in the final report, but instead, are buried in the investigative docket.

Attached to the NTSB Operational Factors Investigator’s Report are 275 pages of interview transcripts with the company’s director of operations, the chief pilot, and several line pilots, which show that other external factors and forces were at work that might have played a role in the crash.

The first is that while both pilots were legally instrument current, they may not have been particularly proficient. It’s not just the pilot’s concerns about her copilot she texted to her boyfriend. Interviews with line pilots reveal that on the same leg a week before, with a different copilot, she asked her partner to fly one of the legs assigned to her because of “how bad the weather looked.”

And her apparent lack of comfort with actual instrument flying was hardly unique.

In interviews with NTSB investigators, company management reported that even though the work of hauling freight happens in the instrument ecosystem, new hires rarely have any instrument time worth mentioning, as most entry level flying jobs — such as flight instruction and banner towing — provide no useful instrument experience.

Applicants have scant experience beyond the minimum required to get the rating; experience tends to be stale, from early in their training; and many have no actual instrument experience at all.

In response to an investigator’s question about what sort of training is most needed in the new hires, the director of operations said, “Overall, basic instrument skills is the number one thing that we maybe do.”

He added that, “they don’t come in with that experience.”

So despite the required higher number of hours to enter the base of the upper echelons of professional flying, the nature of these hours aren’t giving applicants the basic, minimum skill set required for the work.

The second, even darker, aspect that the interviews reveal is the long shadow that the pilot shortage is casting.

The company had a stable team of about 22 pilots throughout the pandemic lockdowns and during the recovery period that followed. But as the airlines ramped up and became increasingly more desperate for flight crews, the company started losing pilots.

In fact, it lost seven of its captains in the six months prior to the accident, practically all of their senior people. Is that why they paired the two new hires together? Because of a lack of senior, experienced hands?

The chief pilot, himself only with the company for three years, said he’d never seen such turnover, saying, “It’s nuts what some of these places are hiring now.”

He related that premium operators that historically required 4,000 hours were now “plucking” first officers with only a year of experience.

“It’s incredible…really incredible,” he said, adding, “it’s been a dog-eat-dog world I feel like lately.”

Were they scheduling rookies to fly together because rookies were all they had? And what does that say for the safety of the National Airspace System?

Has the pilot shortage become a pilot killer?

The Numbers

Want to read more? Download the NTSB’s final report here or view the (many) items on docket here.