Disoriented, Impaired Pilot Causes Fatal Test Flight Crash After Suspected Heart Attack

Something Was Wrong Before the Crash

The Beechcraft Baron was supposed to be performing a simple maintenance test flight.

Instead, almost immediately after takeoff, the pilot began behaving in a way that was inconsistent with his extensive experience.

The airplane was N959CM, a Beechcraft 58 Baron operated by Caribbean World Resorts Ltd.

It departed Ocala International Airport in Florida on October 31, 2019.

There were two people aboard: a 73-year-old private pilot and a 50-year-old mechanic.

Both would die.

The investigation would eventually determine that the aircraft itself had no mechanical problem capable of explaining the loss of control.

Instead, investigators found evidence that the pilot had suffered physiological impairment, although they could not determine whether the impairment resulted from an acute medical event, medication, or both. (NTSB Data)

The ATC recordings make the accident particularly disturbing.

The pilot appeared confused before he was even airborne.

He taxied toward the wrong runway.

After takeoff, he did not respond normally to repeated controller instructions.

He believed he was flying west.

He was actually flying east.

And within approximately two minutes, he asked to return to the airport.

There would not be enough time.

The Maintenance Problem

The flight began because of a maintenance issue.

The day before the accident, the pilot had flown into Ocala and noticed that the right engine's fuel-flow indication was fluctuating.

The indication would move from zero to high values while the other engine indications remained normal.

The pilot asked a mechanic to investigate.

The mechanic removed the fuel-flow transducers from both engines and swapped their positions.

The purpose was diagnostic.

If the problem moved with the transducer, the sensor would be suspect.

If the indication remained with the engine, the problem might be elsewhere in the aircraft's indication system.

The mechanic then performed ground engine runs.

The engines operated without an obvious problem.

The aircraft appeared ready for a local test flight.

Nothing about the maintenance itself suggested that the airplane was about to become uncontrollable. (NTSB Data)

Taxi to the Wrong Runway

At approximately 11:24 a.m., the pilot contacted ground control and requested taxi instructions.

He explained that he wanted to conduct a maintenance test flight.

The controller told him he could choose the area in which he wanted to conduct the flight.

The pilot said he would operate west of the airport.

Ground control cleared him to Runway 18.

But the pilot initially taxied toward Runway 26.

The controller corrected him.

The pilot eventually proceeded toward the correct runway.

At this stage, the controller had already seen something unusual.

The pilot was highly experienced, yet he was having difficulty following a simple taxi instruction.

But the more concerning behavior came after takeoff. (NTSB Data)

The Wrong Direction

The airplane departed Runway 18.

The controller instructed the pilot to make a right turn toward the west.

The controller repeated the instruction.

Again, there was no proper response.

The aircraft turned left instead.

It began flying east.

The controller tried again.

Eventually the controller asked whether the pilot was all right.

The pilot asked the controller to repeat the transmission.

The controller asked where the aircraft was going.

The pilot said he was going west.

The controller had to tell him that he was not.

He was heading east.

That exchange is one of the most revealing parts of the accident.

The pilot apparently did not understand the aircraft's actual direction of travel.

The controller could see the aircraft's position.

The pilot's mental picture did not match reality.

That discrepancy was an early sign that something was seriously wrong.

A Pilot With Thousands of Hours

He held ratings for single-engine land, single-engine seand multiengine land airplanes, as well as an instrument rating.

His logbook showed approximately 7,800 total flight hours.

He had completed a flight review less than a month before the accident.

He held a valid FAA medical certificate.

There was therefore no obvious reason to expect basic navigation or aircraft-control problems during a simple local test flight. (NTSB Data)

That makes the ATC sequence particularly important.

The controller's observations were effectively an early warning that the pilot's cognitive or physiological condition was changing.

The aircraft was not behaving as expected.

And his perception of direction appeared wrong.

The Return Request

The controller asked the pilot about his intentions.

The airplane continued east.

Then, approximately two minutes after takeoff, the pilot transmitted a short request to return to the airport.

It was the final communication from N959CM.

The flight had lasted only a few minutes.

There was no prolonged mechanical emergency.

There was no engine failure requiring a forced landing.

The problem appeared to be inside the cockpit.

The Final Maneuver

GPS data recovered from the wreckage showed that the airplane entered a left turn after takeoff.

It never climbed more than approximately 418 feet.

During the last minute of the flight, groundspeed remained between roughly 95 and 107 knots.

Witnesses saw the airplane flying at low altitude.

Some described it as entering a steep inverted dive.

Others described a left-hand spin.

A video recorded from a vehicle captured the final moments.

The aircraft rotated counterclockwise as it descended toward a four-lane divided highway.

The right wing struck the road first.

The airplane then crossed traffic lanes, struck a curb and came to rest in a vacant area.

A post-impact fire consumed much of the aircraft.

The pilot and mechanic were killed.

A person on the ground was seriously injured. (NTSB Data)

Was the Aircraft Faulty?

The NTSB examined both engines and the fuel-flow transducers.

No mechanical discrepancy was identified that would have prevented normal engine operation.

The engines did not show evidence of a pre-existing failure capable of explaining the accident.

The flight controls were also examined.

Continuity was established for the ailerons, elevator and rudder.

The investigation therefore moved away from a mechanical explanation and toward the pilot's condition. (NTSB Data)

This is important because the flight began as a maintenance test.

The aircraft had just undergone troubleshooting.

It would have been easy to assume that something in the maintenance work caused the crash.

The Medical Evidence

The pilot's autopsy revealed severe abdominal aortic atherosclerotic disease.

The condition increased his risk of an acute cardiovascular or other physiological event.

Toxicology also found diazepam and its metabolites in his tissues and blood.

Diazepam is a sedating benzodiazepine that can impair the ability to perform hazardous tasks.

The toxicology results also detected a very low concentration of THC.

The NTSB concluded that the THC concentration did not indicate recent use and was unlikely to have contributed to the accident. (NTSB Data)

The evidence therefore pointed toward physiological impairment.

But the investigation could not establish exactly what happened inside the cockpit.

Medication may have contributed.

Or both may have been involved.

NTSB Probable Cause

The National Transportation Safety Board determined:

“The pilot’s failure to maintain control of the airplane due to an acute medical event and/or his use of impairing medications.”

— National Transportation Safety Board, ERA20FA022. (NTSB Data)

That wording is deliberately cautious.

The NTSB did not say that the pilot definitely suffered a heart attack.

It did not say that diazepam definitely caused the crash.

That distinction matters.

The ATC Perspective

For an ATC recording sequence, this accident is particularly valuable because the controller's communications show the deterioration before the crash.

The controller was effectively trying to determine why an experienced pilot was not following basic instructions.

First came the wrong taxi route.

Then the missed turn instruction.

Then the incorrect heading.

Then the pilot's statement that he was flying west while actually flying east.

The controller asked if he was all right.

The pilot did not appear to understand the situation.

The aircraft then requested a return.

The radio went silent.

The sequence illustrates a crucial principle of emergency management:

Behavioral abnormalities can be an emergency signal even when the pilot has not yet declared an emergency.

Controllers are trained to recognize unusual aircraft behavior.

But there is a limit to what a controller can do.

The controller can issue headings.

He can provide information.

He can ask questions.

He can coordinate emergency services.

But the pilot must ultimately control the aircraft.

Why the Video Matters

The recovered video provides another independent piece of evidence.

Witness descriptions can be imprecise, particularly during a rapidly developing accident.

The video showed the airplane rotating counterclockwise during the final descent.

It also showed that there was no obvious pre-impact fire or smoke from the aircraft.

That helped investigators evaluate witness reports and distinguish between an apparent engine or fire emergency and the actual sequence.

The video supported the conclusion that the aircraft entered a loss-of-control event rather than simply descending because of an engine failure. (NTSB Data)

Accident Facts

The Aviation Safety Network independently records the accident as a fatal loss-of-control event involving N959CM and identifies the NTSB investigation as ERA20FA022. (Flight Safety Foundation)

The Broader Safety Lesson

This accident is a reminder that pilot incapacitation does not always begin with a dramatic medical emergency.

Sometimes the first warning is subtle.

A pilot may taxi toward the wrong runway.

He may misunderstand a heading.

He may respond incorrectly to simple questions.

He may insist that the aircraft is somewhere it is not.

For controllers, those behaviors can be critical clues.

For pilots, they highlight the importance of recognizing when personal performance is deteriorating.

The tragedy of N959CM is also a reminder that experience does not eliminate physiological risk.

The pilot had thousands of flight hours.

The airplane was mechanically capable of normal flight.

The weather was good.

The maintenance test itself appeared routine.

Yet something happened inside the cockpit that disrupted the pilot's ability to perceive and control the aircraft.

But the safety lesson is clear.

A pilot's physical and cognitive condition is part of the aircraft's airworthiness.

An aircraft can have two healthy engines, functioning controls and clear skies and still become uncontrollable if the person flying it suddenly cannot accurately perceive what the airplane is doing.

For controllers, the final lesson is equally important: when an experienced pilot suddenly behaves unlike an experienced pilot, the abnormal behavior itself may be the emergency.

In Ocala, the controller recognized that something was wrong.

But by the time the airplane requested to return, there were only seconds remaining.

Two people died.

And the investigation ultimately showed that the aircraft had not failed first.

Sources

• Original ATC recording