During 2025, there were a couple of accidents in aviation that made me think. I know I can’t change the world, but I can try.
Just to remind you about a few accidents:
The most recent major aviation accident in South Korea involved Jeju Air flight 7C2216, which crashed at Muan International Airport on December 29, 2024. The plane, a Boeing 737-800, belly-landed, overran the runway, and collided with an embankment, resulting in a fireball. Out of the 179 people on board, only two cabin crew members survived.
Air India Flight 171 was a scheduled passenger flight from Ahmedabad Airport in India to London Gatwick Airport in the United Kingdom that crashed 32 seconds after takeoff at 13:39 IST (08:09 UTC) on June 12, 2025. All but one of the 230 passengers and all 12 crew members died. An additional 19 people were killed, and 67 were seriously injured on the ground.
Germanwings Flight 9525 was a scheduled international passenger flight from Barcelona–El Prat Airport in Spain to Düsseldorf Airport in Germany. The crash was deliberately caused by the first officer, Andreas Lubitz, who had previously been treated for suicidal tendencies and had been declared unfit to work by his doctor.
In all these flights, there was something strange in how pilots handled procedures under extreme stress. In two cases, they switched off perfectly good, working engines. If those two cases were isolated, I wouldn’t even mention it — but let me remind you of another crash in Hawaii several years ago when Transair Flight 810 (a 737-200) lost one engine due to failure, and then the pilots switched off a good working engine and crashed into the sea. There was also an accident at Schiphol Airport when the pilots of a Cityhopper SAAB 2000 set a good working engine to ground idle, then tried to perform a go-around and crashed KLM cityhopper flight 433 on April 4,1994).
In all these cases, there was one common denominator: the investigators could not play back the last few critical minutes of the flight and make 100% sure what caused the crash. Black boxes, although designed to be very sophisticated, didn’t provide enough data to be certain why the accident happened. We may know how it happened, but the why is still a big unknown. Why did the pilot switch off the wrong engine in the Hawaii accident? Why did the Cityhopper pilot pull the throttle to ground idle? Why did the Air India pilot cut the fuel off just seconds after takeoff? And what was the Germanwings pilot doing in the cockpit before the crash?
The whole investigation in all these cases could have been more thorough if there had been video footage from the cockpit. But there was nothing. Nowadays, the technology is available to accomplish that task. There is absolutely no problem installing a few cameras in the cockpit and recording all the actions pilots are taking. All control panels, displays, and pilot actions could be recorded from multiple angles, and nothing would be hidden. Such an installation wouldn’t even be extremely expensive.
Just to remind you — many people use dashcams in their cars. They can record while driving, and in case of an accident, you can see exactly what happened. Such a dashcam can be purchased for less than $50 USD. It is proven that it works fine, and the footage can be used in court.
I was personally involved in several projects related to video recordings. Many years ago, KLM had problems with passengers flying to Suriname and Caribbean destinations. There were regular fights on board, and KLM decided to carry a policeman on board. But within a few months, someone decided it was too expensive to pay for police services. Therefore, they decided to install a camera by the aircraft door to record passengers entering the aircraft and show it in real time on a screen in the cabin. The goal was to create awareness among passengers that they were being recorded — and therefore would not fight. That was, of course, not a good strategy, but management bought it and decided to invest in installing the camera and interfacing it with the onboard video system.
Just a few weeks into the project, the lawyers stopped the activity because of privacy violations. KLM continued flying for another half a year with sky marshals, and then that activity was also discontinued. I never heard of onboard fights after that.
A few years later, the Russian company Transaero hired me to run a project for them. They were having problems with — guess what? Yes, passengers fighting on board. My first reaction was: don’t give them vodka during the flight. But Russians are different. They were afraid that passengers would fly with someone else, so they wanted to keep them by offering vodka. The problem was quite serious — they had fights almost every day, and they had to appear in court every week. The big question was: who delivered the first punch? Therefore, they wanted to have video recording capability on board. And that is exactly what I provided for them.
I installed a server on board capable of recording 24 hours of video from eight cameras. After 24 hours, the system would overwrite the previous recording. So, if there was a fight, after landing, the engineer could simply connect a laptop to the server and download the video. In the beginning, they had three active cameras in the cabin, and later they added two cargo cameras because, at some airports, thieves were opening suitcases and stealing the contents. The video recordings helped identify the crooks.
Eventually, I managed to install such a system in 75 aircraft (all Boeing: 747s, 777s, 767s, and 737s). Transaero went out of business after Putin invaded Crimea and Russia got sanctioned.
Coincidentally, almost 10 years later, I was leading an ARINC committee named Cabin Systems Subcommittee, which standardized such systems for the benefit of aviation. In Denver I even presented that standard — A628 Part 1 (Cabin Surveillance System) — for adoption. In my presentation, I mentioned my project with Transaero, and engineers immediately started asking about privacy. It is obviously a big concern. However, with the Russians (Transaero), privacy was never a point of discussion. Others — especially in Europe — are totally different.
That brings me back to the cockpit and the installation of video recording for monitoring pilot actions. After the crashes described earlier, discussions started about installing cameras and recording cockpit actions. It is no longer a technical discussion. Technically, everything is possible. Besides FDR (Flight Data Recorder) and CVR (Cockpit Voice Recorder), a new box could be introduced: VDR (Video Data Recorder). It wouldn’t be a system for real-time video streaming to the ground, but a much smaller and limited system that can record and store video on board. That’s exactly what we’re already doing with FDRs and CVRs. The data — and the video — should only be retrieved in the case of an accident or incident.
Such systems are already installed in some aircraft like helicopters for search and rescue, and in test and training aircraft. Agencies like the National Transportation Safety Board (NTSB) have often demanded the installation of video recorders. From the investigators’ point of view, it’s obvious — they would have more information, clear visual evidence of the pilots’ actions, and could reach conclusions faster. This discussion has been going on for more than two decades.
But that step is a problem because pilots don’t want it. That’s because of privacy. They are afraid the videos could be leaked to the media or used in court — and what if the video is not used for safety investigations but to assign blame to the pilots? If the videos are made available to the public, they could be misinterpreted and misused. Besides, knowing that video recording is ongoing can influence a pilot’s behavior in a performative way.
It’s obviously a matter of trust. Pilot unions like ALPA, BALPA, and IFALPA are opposing cockpit video recording. They argue that video recording poses a massive privacy risk. They are also afraid that the video could be used for disciplinary purposes.
Let’s deviate a bit here: About 20 years ago, the company Spirent came up with a modification for 747 Classics. They proposed removing the P2 panel (the panel in the middle of the cockpit used for engine instruments) and installing a pallet with displays that fit into the P2 position and used the same connectors as the steam gauges. It was a more reliable system and easy to install. Atlas Air decided to install it in their aging 747 Classic fleet. Technically, it was successful — but the system reported that some pilots were exceeding EGT limits during takeoff. The result was a decrease in engine-on-wing time, which cost the company a lot of money. Now, management and pilots were in conflict. Pilots blamed engineers and computers for providing wrong information, and engineers checked and rechecked the system but found nothing wrong. The bottom line: After some months of operation, the problem disappeared — not because engineers changed the sensor limits, but because pilots started operating more carefully and avoided exceeding limitations.
I’m convinced similar situations could occur with video recordings. I can imagine that pilots would change their behavior simply because they know they’re being monitored. But that change could also have an adverse effect on their performance.
One idea is to set up a system that starts recording in case of a TCAS alert, EGPWS alert or any other warning — and record video for 2–3 minutes. But you wouldn’t know what happened before the alert — and that could be crucial.
Eventually, I am certain it will take a few more years of discussion before video recording becomes a standard cockpit feature. The bottom line is to improve safety, to conduct proper investigations of each accident and incident and to ensure such situations are not repeated. That is in the best interest of pilots, investigators, authorities and the public.