21 AIR Flight 7598 CRASHED Due to PILOT ERROR
People will argue, “You can’t blame the pilots we need to wait for the investigation.” How did that waiting work for the Endeavor Air crash of February 17, 2025? Those pilots are still alive and we heard nothing. I will agree, however, that we cannot blame the pilots if the problem begins with training or a negative safety culture.
When I first heard it was a Prime Air flight, I expected this to be ABX based upon the testimony in the recent Air 21 trial and what I learned in that case. ABX is not training their pilots but for the checkride. Yet, this was 21 Air. Why did they crash? Was it due to lack of experience on the B767? Or due negative safety culture?
Pilot Experience Breakdown:
- Captain Joseph Carroll (Age 55):
- Total flight experience: 7,145 hours
- Boeing 767 type rating obtained: May 2026
- First Officer Jaime Felipe Silva Molina (Age 37)
- Total flight experience: 2,655 hours
- Boeing 767 type rating obtained: April 2025
What is pointedly missing from the NTSB report of flight hours is their time in the 767. Both pilots were new to the aircraft. They both had limited flight time, and if not trained properly what else could we expect?
CHILLING DATA
7 minutes, 35 seconds 30 seconds before the end of the recording, the nose landing gear and the right main landing gear touched down and ground speed was 158 knots.
7 minutes, 49 seconds 23 seconds before the end of the recording, brakes were applied, ground speed was 146 knots.
7 minutes, 58 seconds 19 seconds before the end of the recording, the left main landing gear touched down at a ground speed of 134 knots.
8 minutes, 9 seconds 15 seconds before the end of recording, brakes were released at a ground speed of 120 knots.
8 minutes, 17 seconds Throttles increased to values consistent with go around thrust. 11 seconds before the end of recording.
8 minutes, 25 seconds Throttles were reduced to idle power.
8 minutes, 28 seconds Brakes were reapplied and the ground speed was 117 knots. The throttles remained at idle and brakes continued to be applied until the end of the recording. 7 seconds before the end of recording.
8 minutes, 42 seconds Lateral, longitudinal, and at vertical accelerations changed.
8 minutes, 48 seconds Ground speed was one uh was 96 knots at this time and 0 seconds at the end of recording or the last data frame. Final ground speed was 65 knots.
END.
WHO was flying this plane? Maybe both. Maybe nobody. After listening to these details, the pilots were clearly confused as what to do … Stop? Go? And when you’re touching down on a runway 3000 feet beyond the marker at 158 knots, you might have a problem to begin with. No reverse thrust means they either forgot or were not committed to the landing. The cockpit voice recorder will be fascinating, I hope they don’t hide that data as they did with the Endeavor crash.
Listen to the NTSB MEDIA BRIEFING TWO here.
The NTSB said, “Tomorrow we have interviews scheduled with the assistant chief pilot and both pilots of this aircraft. We will also be interviewing the chief pilot for 21 air, but that has not been scheduled. It will be scheduled.” Waiting for that interview.
Why didn’t the NTSB talk to the Endeavor pilots? If they did, why didn’t we hear about it? Both flight crews on each plane crashed their aircraft because they failed to go around during an unstable approach. In 21 Air’s case is appears they were completely out of control on the runway with indecisiveness as what to do. But any pilot would know that if the wheels were not on the ground by 3000 feet at 158 knots (I think they should have been about 145 kts) they should have got out of there.
Could we have learned from Endeavor crash if the FAA held Delta accountable? Probably.
Read Learning from Accidents Here.
If the transcripts to Endeavor crash had been released and interviews with the pilots had been conducted and the FAA decided to hold airlines accountable to their training and duty time regulatory failures, and mandate all airlines to become compliant, then maybe this accident would not have happened. The irony of the airline that crashed is called 2`1 Air, and the Whistle blower law is Air 21.
Yes, 21 AIR has an Air 21 complaint against it with Seuring v. 21 Air, as did ABX (also carrying cargo for Prime Air) and there were AIR 21 complaints against Delta and Endeavor, prior to the Endeavor crash. More on that coming soon.
The Miami Herald did not quite get the facts correct with the Seuring case. I’ll explain why in the next post. But these Air 21 cases, if the courts had any administrative power to impose sanctions, might be the solution to future airline safety.
21 Comments
On another thread I challenged a comment suggesting a deep dive into ‘background causes’ so that “it never happens again”. What a hope!
These handling accidents have been happening since 1965 to my certain knowledge. In reality they are less and less excusable… instrumentation and procedures are much more refined. Stable approach criteria have been introduced and still these crazy accidents happen. The fact is that more and more procedures reduce frequency but don’t prevent.
I have always been an advocate of sacking pilots who have more than one landing off unstable approaches. Pour encourager les autres,
What is so special about pilots that we seek reasons for misbehaviour? Sometimes the just culture is too just.
This appears to be gross negligence and/or incompetence from the cockpit. There is also strong evidence that crew fatigue was a contributing factor. I wonder if this will be the end of the “cargo carve out” from Part 117 pilot fatigue rules?
Karlene Petitt PhD, MBA, MHS I have a feeling there will be renewed interest in the Safe Skies Act that will merry up fatigue rules between cargo and passenger after this incident.
I believe the root cause lies elsewhere. It’s merely a reflection of our society—the absence of common sense, the neglect of attention to detail, and the lack of consequences for bad actions. For instance, you can fire someone simply because they’re not good at their job, and your actions are publicly criticized on social media, making you the villain.
Today, I witnessed several unusual traffic light incidents. Despite the red light, five cars continued rolling through the intersection. I also saw a Tesla auto driving while its driver was scrolling his phone with both hands.
And now pilots glided across half the runway at high speed before attempting to stop on the remaining half.
No matter how much training you receive, if you don’t take a moment to think and apply common sense, you’re doomed. Look at the poor pilots of Air 21; they’re already facing the “justice” of social media. How terrible they are—these individuals are criminals and deserve to be in jail. All those “Wikipedia Professor” are giving their opinions on something they don’t understand. Some of them have probably never even set foot in an airplane.
Ok I take a pill and go to bed 🙂
I have been waiting to read the Cockpit Voice Recorder for quite sometime now of the Endeavor Air Toronto RJ accident. Why the secrecy?Would suggest something to hide perhaps? We are supposed to learn in aviation about other’s mistakes. We were lucky on 21 Air they carried no passengers or this accident may have been way worse.
When did full power and the nose above the horizon ever make an unstable approach worse?
It’s not about decision making, communication and nine core competencies and all the intellectual clap trap that smothers aviation.
It’s a binary decision yes or no, if two intelligent well trained pilots can’t work that out in the seven minutes they spend on an approach then something in training procedures is lacking.
All that time to decide “this isn’t working” is child’s play and to complicate it unnecessarily will not cure the problem.
Yes, Karlene, training and professionalism is essential across commercial flying. Another incident similar to this Amazon/21 Air is the Southwest 1455 overrun at BUR and attributed to excessive speed on final with no go-around.
NTSB report: https://www.ntsb.gov/investigations/Pages/DCA00MA030.aspx
It’s all related to a situation when CRM fails called : “Control Takeover Training” ..the First Officer expressed concern about approach speed ..he was not listened and with the tunnel vision by the captain who was already committed to land ..the FO failed to take controls and execute a go around/rejected landing . FOs are not usually encouraged or trained to take over controls based on their judgment based on flight safety. No operator minds about this.
These recent things seem to be lack of basic flying skills. Pilot training teaches 1) how to fly a generic single engine piston trainer and 2) how to fly a specific model of transport category jet (with the autopilot and autothrottles on all the time) but not how to fly a generic transport category jet without automation.
And people thought I was off my rocker last month when I said the CVG tail scrape could have very easily been this same crash in Miami. Unstable means go around…….period. Manual fire wall the engines. Toga. Flaps 20. Retrim aircraft to rotate and climb out at VAC. Verify ground spoilers retracted even though they should stow with thrust levers coming up. This is basic airmanship. Basic. The people running these companies could care less about civilian casualties. They will have a replacement aircraft in very short order. People are NEVER replaceable. My thoughts and prayers remain with those affected by this unfortunate tragedy
The answer is simple Karlene Petitt PhD, MBA, MHS . Think of the pipeline to ATP that existed when you were a student pilot. We gained REAL EXPERIENCE instructing or flying freight or checks before a regional airline job. That pathway is generally not used anymore because the 1500 hour rule has made flight time and a type rating the priority. Freight companies like 21 Air see new FOs with 1500 hours and an ATP with a brand new type rating in a B737 or A320 as hireable. In general they have no jet experience at all and likely no experience in general depending on the QUALITY of their 1500 hours the logged. This assumption that they can fly a complex jet aircraft with little to no jet experience is what the problem is. Until the FAA changes its certificate and testing requirements to require SBT training and checking and real world experience we will continue to have a lack of critical thinking on the flight deck. The advent of AI has made this worse in an aviation culture that already took shortcuts to study and train to “pass the exam” and giving no practical experience.
All the points made here are real and valid.
However, the “root cause” will be a messy tangle of many interactions, all and none of which will be the “smoking gun” which resulted in this accident.
Let’s focus on what we can directly and tangibly influence: training.
Currently we aim to the lowest level of acceptable pilot performance. Let’s change our goal from this to the aim of developing lifelong expertise.
I believe there was an AIR21 case involving 21 Air in the 2020-2023 timeframe. I flew with (former airline pre 9/11) the then Chief Pilot who was trying to sound the alarm and was fired for his advocacy of safety. 21 Air has had issues for some time apparently and the FAA allowed them to continue on.
Hi Karlene, I think you gave me a 757 simulator ride at AWA around 1991. Here is one suggestion: The FAA could appoint, or designate retiring Part 121 Captains as Check pilots on the aircraft they retire on. Trained and certified by the FAA for that specific purpose, their experience and expertise could be passed on during these events. Acting independently, under the FAA umbrella, there would be zero reason for any bias towards allowing marginal(or less) pilots to continue their progression without intervention. Simulator events as well as line observations. The FAA has several Voluntary Safety Programs, including FOQA. Make those mandatory for Part 121 operators. The IEP, Internal Evaluation Program consists of self auditing programs which may lead to Voluntary Disclosure reports per the VDRP. Financial and airline staffing requirements may reduce an operator’s propensity to report. If the FAA cannot staff their ranks for oversight, put independent, experienced professionals on the job.
A recent article in an on-line newsletter ( Why 97% Of Boeing 777 Pilots Continue An Unstable Approach Instead Of Going Around) indicated through flight data monitoring, that most unstable approaches in Boing 777s result in the crews continuing the approach to landing, rather than executing a go-around. In fact, 97% of the approach data gathered that qualified as an unstable approach indicated that the crew continued the approach to a landing. The good news is that unstable approaches examined account for only 3.5-4% of all approach data examined. If that sounds like a lot of approaches going “wrong” we could all argue you are correct. That said, the article cites the Flight Safety Foundation data indicating that the organization developed the definitions and criteria for a stabilized approach “after its own research found that unstabilized approaches were a causal factor in 66% of 76 accidents and serious incidents during the final stages of those flights between 1984 and 1997.”
This is a cultural issue with pilots thinking they can “save” a bad approach.
There are many steps along the way to get to where two pilots, flying into their home domicile can screw up this bad. The pilots themselves, background, experience? Company culture, pressure to save time? Salary, rock bottom I’ll bet. Start up carrier to make a buck hauling boxes. Training for type? Who administers? FAA oversight or lack there of? FOQUA data use? Line orientation training? Green on green rules? Normalization of deviance on unstable approaches? I predict the company will file Chapter 7. Good riddance.
I don’t know if it has already been pointed out, but this is a basically a redo of the 2005 Southwest flight into the Midway airport.
There were weather factors in that event as well which compounds bad decisions.
We all know a bad landing starts at least five miles and often further, when the basic decision process is established as to how things will proceed.
I don’t know if there are penalties from Amazon should a crew not make their time window, or if there are other pressures motivating a crew to not say to the controller that they are not in a stable position and would like to go back to the end of the line to get better setup.
What ever the underlying problem was in this event, I am sure the human factors and the external or self imposed pressures will surface
Until management isn’t part of the ones that sit on the court this won’t change….does the term accountable manager means anything???? Not really!
Who hires the pilots?
Who supervises the pilots?
Who’s responsible for the quality of training?
It is very easy to blame the pilots!
Powerful points, Karlene. Comparing these incidents highlights how critical proactive training and regulatory oversight are before tragic outcomes happen. Thank you for pushing this important conversation forward.
Why are you people still talking about this nothing will ever get fixed
See this is the problem, everyone talks about it, but doesn’t do no good, cause your talking to the wrong people. People need to stop flying and after a month of billions lost then they will finally do something. No one needs to fly in 2026, everything can be done on zoom, internet or telephone, and no one needs to fly to vacations especially as expensive as everything is now a days and no one can afford it anyways. People need to start using common sense or better yet their brains.