Surveillance Cameras for Pilots

[anorlunda]

Most of us have cars with backup cameras. They are inexpensive, simple, lightweight, and reliable. Teslas have as many as 7 cameras looking in all directions.

I think that cameras showing views of wings, stabilizers, rudder, engines, landing gear, cabin, cargo hold, would be of great use to pilots in flight. Why don't they have that available in commercial airliners? Just think of cases where smoke is reported in the cabin or cockpit, or where pilots experience flight control problems, or fuel leaks as in the Air Transat 236 incident.

In the modern world, we also see more and more use of image analysis used cleverly. Engine compressor stall is something that apparently is hard to see from the conventional instruments, yet it is immediately obviously to passengers looking out the windows. A video analysis could provide a better compressor stall warning. A forward looking camera might be able to signal "possible bird strike detected".

I do understand that retrofits can be very expensive. But I haven't heard about cameras even in the newest model airliners.

 

[jedishrfu]

Apparently, they are doing it already, although quietly.

https://www.businessinsider.com/why-airplanes-have-cameras-what-they-do-where-to-find-2022-1

 

[anorlunda]

Apparently, they are doing it already, although quietly.

https://www.businessinsider.com/why-airplanes-have-cameras-what-they-do-where-to-find-2022-1

That's cool. So the principle is there, and being used. Now they just need more. But it sounds haphazard. One camera and monitor for cockpit door few, other systems plus other monitors for different views. No unified system able to switch between hundreds of views.

That view from the top of the tail shows the wings and the engines. That view alone could have played a role in numerous past airplane incidents.

The most emotional one to me was the Air Transat 236 incident. The plane has redundancy, fuel tanks both port and starboard. One side suffered a massive leak. The pilot, after studying the information available to him, decided to open the transfer valve, thus defeating the redundancy. The result was out of fuel on both sides, while flying over the Atlantic. If only he had a better way to see the fuel leaking.

 

[jedishrfu]

Flying is always fraught with danger.

The Air France disaster where they chose to fly around one storm only to run into another. The water vapor froze on the pitot tubes causing the automation system to think the plane was stalling with disastrous results.

https://en.wikipedia.org/wiki/Air_France_Flight_447

NOVA did a great documentary on it and the effects of supercooled water.

https://www.pbs.org/wgbh/nova/space...On June 1, 2009, Flight,loss of all 228 lives.

 

[russ_watters]

That view from the top of the tail shows the wings and the engines. That view alone could have played a role in numerous past airplane incidents.

I think they could help if it's cheap, easy and light, but usually I'd think the data is more important than the view.

The most emotional one to me was the Air Transat 236 incident. The plane has redundancy, fuel tanks both port and starboard. One side suffered a massive leak. The pilot, after studying the information available to him, decided to open the transfer valve, thus defeating the redundancy. The result was out of fuel on both sides, while flying over the Atlantic. If only he had a better way to see the fuel leaking.

That doesn't make much sense to me, and the pilots did get some blame for failing to understand from the data that there was a fuel leak. Maybe cameras would have helped, but they would have had to have a reason to check.

 

[anorlunda]

That doesn't make much sense to me, and the pilots did get some blame for failing to understand from the data that there was a fuel leak. Maybe cameras would have helped, but they would have had to have a reason to check.

I can't disagree because we can't get inside the pilots head. Here's some relevant passages from the report that I think are interesting.

At approximately 05:33, an advisory ADV message was displayed on the Engine/Warning Display (EW/D). The crew noticed the ADV and deselected the ENGINE ECAM page. This action resulted in the Fuel ECAM page being displayed and the crew becoming aware of a fuel imbalance between the left and right inner-wing tanks. To correct the imbalance, the crew selected the cross feed valve OPEN and the right-wing fuel pumps OFF in order to feed the right engine from the left-wing tanks.

...

8. The flight crew did not detect that a fuel problem existed until the Fuel ADV advisory was displayed and the fuel imbalance was noted on the Fuel ECAM page.
9. The crew did not correctly evaluate the situation before taking action.
10. The flight crew did not recognize that a fuel leak situation existed and carried out the fuel imbalance procedure from memory, which resulted in the fuel from the left tanks being fed to the leak in the right engine.
11. Conducting the FUEL IMBALANCE procedure by memory negated the defence of the Caution note in the FUEL IMBALANCE checklist that may have caused the crew to consider timely actioning of the FUEL LEAK procedure.
12. Although there were a number of other indications that a significant fuel loss was occurring, the crew did not conclude that a fuel leak situation existed – not actioning the FUEL LEAK procedure was the key factor that led to the fuel exhaustion.

..

The [FUEL IMBALANCE] procedure also contains a CAUTION note that, if a fuel leak is suspected, the crew should instead refer to the FUEL LEAK procedures.

...

A FUEL LEAK is an abnormal condition that is not monitored by the ECAM system; consequently, an ECAM warning is not generated for these conditions. Instead it is expected that the crew’s normal monitoring of the fuel system parameters will result in timely recognition that a fuel leak exists or a loss is taking place, and the conclusion that actioning of the FUEL LEAK procedure is required.

By "normal monitoring", I guess that they mean comparing estimates of fuel remaining with instrument readings of fuel remaining.

11. The lack of training in the symptoms of fuel leak situations resulted in this crew not being adequately prepared for the situation encountered on the occurrence flight.

So they blame a lack of training. But the CAUTION note on their placard, does not give hints on how to tell the difference between an imbalance and a leak.

If I were the pilot, I would want concrete evidence of no leak before opening that valve. If video was available, I would check it.

There is a weak analogy between this and the redundant AOA sensors on the 737MAX incidents that we discussed so much. Both cases have to do with handling of redundant resources.

 

[.Scott]

That view from the top of the tail shows the wings and the engines. That view alone could have played a role in numerous past airplane incidents.

The most emotional one to me was the Air Transat 236 incident. The plane has redundancy, fuel tanks both port and starboard.

That leak was in a fuel line - even if it coated a piece of wing surface came in view from the tail cam, it would take probably take direct sunlight for it could be spotted. I have seen the tail cam video feed as a passenger in cross-Pacific flights.

Flying is always fraught with danger.

The Air France disaster where they chose to fly around one storm only to run into another. The water vapor froze on the pitot tubes causing the automation system to think the plane was stalling with disastrous results.

https://en.wikipedia.org/wiki/Air_France_Flight_447

The AF447 autopilot system did not stall the plane. It decided it didn't have reliable information and turned the flying over to the pilots. The pilots had not been allowed by company policy to do real (non simulator) flying - so they muffed it, stalled the plane, and didn't recognize the stall until it was too late.
The crew probably already knew that some airspeed reports were inaccurate, but perhaps the image of an iced-over pitot tube may have made things click for them. But the three Airbus 330 pitot tubes are located on the lower half of the fuselage near the nose. It would be tough to position a camera that would catch an iced-over pitot tube without becoming thoroughly iced-over itself.

 

[anorlunda]

That leak was in a fuel line - even if it coated a piece of wing surface came in view from the tail cam, it would take probably take direct sunlight for it could be spotted.

That's probably true. Any camera view may or may not be useful in specific circumstances.

My argument is that the costs are low, so even if the probability of utility is less than 50%, it would still be worth while.

 

[.Scott]

When looking for an airline accident that may have been avoided with CC cameras:

ValuJet 592 comes to mind. A few cameras in the cargo area may have convinced the pilot to stay on the ground.

Camera in the landing gear compartments may reduce the incidents of stowaways falling from the sky.

 

[Flyboy]

Aviation tries to lean on the KISS principle whenever practical. Yes, there are absolutely times where such cameras would be useful. But, what is the value outside of those rare cases? Is it worth the added complexity?

Camera(s) positioned so that pilots can see the engines could be useful in certain scenarios, but the view angles that would be most helpful are, unfortunately, rather hard to get: Out in front of the engine, looking down the inlet.

Wings, tails, fuselage, etc., cameras would need some standoff to provide the required perspective to properly spot leaks.

As an aircraft mechanic, I really don't think they're worth much. Better sensors and sensor placement feeding into the central computer honestly is a much better investment, as they can provide much more data under a wider range of situations, especially if paired with a flight data recorder that tracks those sensor outputs. That data can be extremely helpful in diagnosing problems, especially if the crew makes note of what time an issue cropped up so you can scroll to that particular point in the flight and scan through the data there.

 

[DaveC426913]

I think the thing about sensors is that, by and large, the data of a single sensor is two dimensional - it records one parameter over time. If anything goes awry, that single parameter might stop making sense (I mean, it could just stop recording correctly. Does a single sensor have any error checking?) - at least until it's put in the context of other sensors.

A camera's data is so rich it records data in three - even four dimensions. (3D physical coverage over time). It's highly nuanced, and self-checking (in the sense of knowing if its recording valid data or busted.

The sensor reading on the deployment of a control surface might be confusing if it's just a single variable over time, but a video of that surface might show the surface deforming or failing.

I also think I don't know what I'm talking about when it comes to the sophistication of aircraft sensors, so this is really more of a Devil's Advocate position. An opportunity for some smart people to educate us about sensors.

 

[Flyboy]

For your example of the control surface... if you’re using a computer to process the data, you can program it with expected sensor ranges during normal operation and anything outside that limit can trigger an alert. Usually any tendency to deform is well understood by engineers, as if it’s too sensitive to that, you get “flutter”, which can and will cause prompt fatigue failure or even catastrophic structural failure.

Yes, sensors like temperature or pressure or volatile hydrocarbon vapor sensors do only provide whatever data they’re designed to output, and can be as simple as a binary output, like a switch. Without context, the data is junk. But with context, with understanding of what that output represents, it’s tremendously helpful. Hell, even the data from a failing or failed sensor can be useful in understanding why it failed, and by extension, what else is wrong.

Case in point: there’s a single float-driven potentiometer in each fuel tank in a Cessna 172 wing. 26.5 gallons of total fuel in each tank. Pretty bog standard stuff. Float moves, signal changes, gauge or indicator on MFD changes accordingly. Simple enough, right?

What does a wildly fluctuating needle on the fuel gauge mean if you’re sitting still on level ground? You might think it’s a bad gauge, or a bad sensor. But in this case, it’s a known “bug”, for lack of a better term, with the sensor design dating back 60 years. When full to the brim, the potentiometer is in a position where it effectively glitches and tries to indicate both full and empty at the same time. But burn a pound of gas off, and the problem goes away.

 

[DaveE]

It is easy for us sitting in our comfy chairs in front of our computers to say "they should do this...". IRL there are a lot of practical limitations. Here are a few things to consider:

1) Commercial aircraft have an extremely tough reliability requirement. Just because Google can do something impressive doesn't mean you want it in the aircraft fleet. I designed stuff for satellites many decades ago and quickly learned that Hi-Rel = Old Tech. Proven reliability is important. If something can't fail, let someone else use it first. If you have something that works, don't be adding extra stuff to it without a lot of analysis.

2) Commercial airlines are in an extremely competitive business, they are doing well just to stay solvent. They aren't going to be spending money on "a better way" unless it improves their P&L statement. Air travel is already extremely safe, and most accidents aren't going to be fixed with a simple technology addition. From a safety perspective, training the humans (mostly pilots) to do what they are supposed to is where the biggest payoff is.

3) Commercial airlines operate a complex systems that includes aircraft technology and all of the necessary support; mechanics, pilots, etc. You can't just toss in a new bit without developing lots of support; procedures and training for everyone affected. It's not just a camera, it's an extra thing for the flight crew to deal with under stressful conditions. Maybe it would help, but maybe it would be a cause of unnecessary distraction, like Eastern flight 401 that slowly crashed into the everglades because the entire flight crew was focused on a landing gear indicator problem. Look at the (poorly implemented) MCAS system added to the 737 Max fleet just to avoid making pilots get a different type rating because airline wouldn't want to spend the money required to retrain pilots.

4) Regulatory agencies are involved. You can't just add stuff to an airplane without talking to the FAA et. al.

So, I agree this would be a good thing, as do the aircraft manufacturers. Some already have cameras, a little bit. They are working on it, but they will need time to implement it in the big airplanes. Things move slowly in that industry, for good reasons.

 

[Filip Larsen]

It is also hard to imagine how camera should be incorporated in the heavy procedure-oriented workflow in the cockpit. Modern work flow for abnormal situations is that the avionics (sensors, computers, displays) report an issue and perhaps even presents a check-list, and the pilot then proceed with that list. How do cameras fit into this? When is the pilot supposed to use a camera to inspect a piece of equipment and what should he look for?

 

[Rive]

But I haven't heard about cameras even in the newest model airliners.

I think you are just looking at the issue in the wrong angle. As something for the pilot, well ... based on the safety and reliability of planes it's not really needed, and by the usual attitude of authorities it would cost a fortune to be recognized as an instrument for flight.

By my humble opinion, it'll come first as a passenger thing. Likely not that extreme as on that link, but many planes already has some kind of personal displays, and to have a live feed of choice from some cameras around is just matter of time.

 

[DaveC426913]

It is also hard to imagine how camera should be incorporated in the heavy procedure-oriented workflow in the cockpit.

I was assuming cameras would only be referenced in a maintenance/reconstruction environment. It's overkill for the pilots.

 

[russ_watters]

It is also hard to imagine how camera should be incorporated in the heavy procedure-oriented workflow in the cockpit. Modern work flow for abnormal situations is that the avionics (sensors, computers, displays) report an issue and perhaps even presents a check-list, and the pilot then proceed with that list. How do cameras fit into this? When is the pilot supposed to use a camera to inspect a piece of equipment and what should he look for?

I would think the cameras would be there for emergencies only. There are situations that call for a pilot to walk back through the cabin and inspect the wings/engines through the window (im not sure if it is actually on an emergency checklist). A camera would be an adequate substitute for that.

 

[Flyboy]

I would think the cameras would be there for emergencies only. There are situations that call for a pilot to walk back through the cabin and inspect the wings/engines through the window (im not sure if it is actually on an emergency checklist). A camera would be an adequate substitute for that.

On airliners, you can usually use cabin crew for that, allowing both pilots to remain on the flight deck to help manage the situation. Might be different on freighters, though, as A: No cabin crew, and B: no windows.

 

[Vanadium 50]

Consider UA232. Your camera tells you you've lost hydraulics.

OK, now what? How does that help the pilots?

 

[Flyboy]

Consider UA232. Your camera tells you you've lost hydraulics.

OK, now what? How does that help the pilots?

An excellent example. All a camera could do is confirm the extent of the damage from the engine coming apart. You'd be able to get most of the same indications from the flight engineer's station on the DC-10, but it wouldn't help the crew figure out a way to save the aircraft any better.

To be honest, there wasn't anything that visual confirmation could have done to improve the odds of survival. They did everything possible to save lives, and it worked. There were survivors from a scenario where, realistically, no one expected anyone to survive. And that single accident helped reshape the way that aircrew can respond under those kinds of conditions. Hell, it inspired more than a few research programs into alternate means of control by adaptive networking and flight computers accommodating and accounting for damage.

That's where the investment should be getting made: resilience to damage, helping crew maintain control of the aircraft in an emergency, and giving the crews better training on how to react under emergency scenarios, not to diagnosing what's wrong visually by adding cameras. Along with better training of crews to understand how the plane works, systems-wise, to help their ability to adapt to a crisis. The aforementioned fuel loss incident upthread? The crew could have noticed that the crossfeed wasn't resulting in the fuel levels on the other side going up, and shut off the crossfeed pumps and contained the issue. UA232 did notice their issue, and rapidly switched to triage and finding an alternate means of controlling the plane. The Miracle on the Hudson? Same thing. They wasted no time trying to figure out what went wrong, or how bad the damage was. They immediately jumped into maintaining control of the plane and assessing their options. No time spent faffing about with MFD and EICAS menus to get more info, just dealing with the crisis at hand.

On the other hand, if you look at the fatal accidents in air transport over the last 40 years or more, many of them can be laid at the feet of air crew not responding to the situation correctly, either due to incorrect procedures, or simple inexperience with the kind of scenario they found themselves in. Air France 447? The crew lost cohesion and communication under the crisis, they didn't realize the change in how the plane's flight computer was interpreting inputs from the controls, and lacked adequate training to recognize the situation they were in and how to correctly respond to the stall.

The list of incidents that could have been fatal but weren't? Most of them had the crew responding to the situation correctly, due to a combination of accurate information provided by the aircraft systems and manuals, and crew experience.

It really boils down to the non-physics topic of "human factors", which is a can of worm I don't want to crack open here.

 

[.Scott]

I've looked at all of the fatal US Commercial Plane accidents from the last 25 years with enclosed cockpits - or large enough for an enclosed cockpit.

It doesn't really show any compelling reason for cameras. But some are interesting.

2022 Fatal Commercial Plane Accidents
From that list, I only include planes with enclosed cockpits:

Wiki Descriptions	Would a camera have helped.	Notes.
An airline worker was pulled into the engine of the parked aircraft and killed.	No	The pilot had full visibility of the worker at the time.
A man who had illegally entered the airfield was struck by the plane as it was landing.	No	Already visible to pilot.
The aircraft overshot the runway during landing, killing one passenger.	No	Pilot error. Nothing unknown to see.
The cargo aircraft crashed during final approach.	No	NTSB sites pilot error, "spatial disorientation"
A Horizon Air employee stole an aircraft and committed suicide by intentionally crashing it over an hour later.	Not likely.
The aircraft experienced a contained engine failure with debris penetrating the fuselage; one passenger was partially ejected from the aircraft and later died of her injuries.	No	There would have been no predictive indications visible to a camera. There were immediate and ample indications of the engine failure and cabin decompression. The pilot responded by immediately returning to the airport.
The cargo aircraft crashed short of the runway during landing because of pilot error.	No	The pilots had ample warning that they were low but continued their approach to landing. Attempting to save a bad approach (instead of using a go-around) is well-known as a bad idea.
The aircraft crashed short of the runway during landing because of pilot error.	No	NTSB sites cockpit crew error, "mismanagement of the airplane's final approach"
The cargo aircraft crashed into a mountain during approach because of pilot error.	No	This was "controlled flight into terrain under instrument conditions".
The aircraft crashed into a house during approach due to the flight crew's improper response to an impending stall, killing all on board and one in the house.	No	Not unless the camera was pointing to the aircraft flight manual.
The aircraft mistakenly attempted to take off using an incorrect runway that was too short, causing it to overrun the runway and crash.	No
The aircraft experienced in-flight structural failure caused by metal fatigue cracking and poor maintenance and crashed, killing everyone on board.	No	Literally, a wing suddenly ripped off. I've listed this because I believe this aircraft can be configured with a closed cabin - but it's getting close to the limit.
The aircraft overran the runway at Chicago-Midway during a snowstorm because reverse thrust was not applied in a timely manner. Everyone aboard survived, but the plane crashed into a vehicle outside the airport and killed a child in the car.	No
The aircraft crashed on approach because of pilot error.	No	They were looking into the night darkness for the airport while they should have been checking their altimeter.
The ferry flight suffered dual-engine failure at high altitude and a subsequent crash caused by the crew's reckless flying.	No, not really.	It was the crew's last "yeee-haaa" moment. But I'm sure they could have found some entertaining use for camera's.
Mismanagement of fuel tanks caused a dual-engine flameout, and the cargo aircraft then crashed during approach, killing the first officer.	No	NTSB: fuel starvation resulting from the captain's decision not to follow approved fuel crossfeed procedures. "Fuel starvation" is when there is fuel in the tanks but none reaching the engines. "Fuel exhaustion" is when you run the plane dry.
Because the crew did not perform a preflight checklist, faulty maintenance on the trim system was not detected and the ferry flight crashed shortly after takeoff.	No	A "ferry flight" is one to transport the plane itself - often for repairs.
Overweight and with faulty maintenance to its elevator system, the plane entered an unrecoverable stall just after takeoff.	No
The aircraft lost its vertical stabilizer shortly after takeoff as a result of the first officer's extreme rudder inputs during wake turbulence and crashed into a neighborhood in Queens, killing all on board and five on the ground.	No	The wake turbulence from from a Boeing 747 that had departed a minute and a half earlier.
[ The four aircraft hijacked on September 11, 2019. ]	Possibly.	If cameras had provided warnings to the crew of an attack, things could have good very differently. Even with the changes that have been made since then (reinforced / locked cockpit doors), cameras might still be useful.
In an apparent case of air rage, a man opened the cockpit door and was then subdued by eight others and died of asphyxiation as a result.	Possibly. And if the camera was recording, yes.	A properly positioned camera in the passenger compartment may have allowed them to see the situation develop sooner. It would also have allowed the cockpit crew to monitor the situation as it evolved. If the camera had been recording, it may have provided useful legal evidence.
The cargo aircraft crashed shortly after takeoff after the right elevator control tab detached because of faulty maintenance, causing loss of pitch control.	Maybe. But not likely.	A camera in just the right spot would have seen an elevator control rod detach during rotation ("lift off"). The pilots initial diagnosis of the problem was a possible weight shift in the cargo. So a camera on the cargo might have downplayed that possibility. It either case, the additional information would not have likely saved the flight - and the location of the crash did not result in death or major damage to property on the ground.
The aircraft lost pitch control when an inadequately lubricated jackscrew tore loose, causing it to crash into the ocean while preparing for an emergency landing.	Possibly. But not likely.	A camera in just the right spot in the wings interior may have provided enough information and/or earlier information about the mechanical problems.
The aircraft crashed into the ocean because of deliberate flight-control inputs by the relief first officer, though Egyptian authorities maintain that there was an underlying mechanical cause.	No.
After the crew opted not to divert to another airport in inclement weather, the aircraft overran the runway when the spoilers were not deployed upon landing.	No.	From wiki: The pilots rushed to land as soon as possible, leading to errors in judgment that included the crew's failure to complete the airline's pre-landing checklist before descending.
Improperly loaded cargo resulted in faulty stabilizer trim settings that caused the aircraft to pitch over uncontrollably, stall and crash just after takeoff.	No.	The aircraft was out of trim and out of its weights and balance limits.
The aircraft experienced a controlled flight into terrain due to pilot error, air traffic control error, and inclement weather. The ATC Minimum Safe Altitude Warning (MSAW) system was deliberately modified so as to limit spurious alarms and could not detect an approaching aircraft that was below minimum safe altitude.	No.	A couple of failures by crew and others to follow procedures.

 

[berkeman]

It doesn't really show any compelling reason for cameras.

So you're basically saying, "Nothing to see here folks, move along..."

 

[DaveC426913]

" the additional information would not have likely saved the flight..."

Again, I wasn't thinking of in-flight use so much as flight reconstruction.

But now that I think of it, cameras produce a LOT of data. Would the additional systems required to record all that data be prohibitive? Would it all have to get stored in the Black Box to be recoverable? I'm now reconsidering whether the benefit of cameras outweigh the cost.

 

[Flyboy]

But now that I think of it, cameras produce a LOT of data. Would the additional systems required to record all that data be prohibitive? Would it all have to get stored in the Black Box to be recoverable? I'm now reconsidering whether the benefit of cameras outweigh the cost.

There would be a lot more storage required, even with modern solid state storage. I don't think it's worth much, especially given how heavily instrumented airplanes are these days.

 

[DaveC426913]

There would be a lot more storage required, even with modern solid state storage..

Yeah, although there's ways around that. My dashcam records continually but only keeps the last 45 minutes. That's A $40 SD chip.

During a Commercial flight, there's little cause to keep video from, say, four hours ago.

But I recognize it's not as simple as that. A camera with an SD chip wouldn't be found in a crash scene like a black box would.

 

[MiaisMia]

It is also hard to imagine how camera should be incorporated in the heavy procedure-oriented workflow in the cockpit. Modern work flow for abnormal situations is that the avionics (sensors, computers, displays) report an issue and perhaps even presents a check-list, and the pilot then proceed with that list. How do cameras fit into this? When is the pilot supposed to use a camera to inspect a piece of equipment and what should he look for?

I think that "the camera" should be treated like "the black box". It's not there to help the pilot look at things, just as the black box doesn't help a pilot to go back and listen to things. It is there in the event of a situation or accident that needs review and to help prevent such an incident from ever happening again. Pilots do not have to incorporate any procedures on their part to make sure the box is doing what it should be doing. Same idea with the camera...automatic.